# morph: sources

> every source cited in the morph history, with its date and type.

- [aider edit-format leaderboard](https://aider.chat/docs/leaderboards/edit.html) — aider.chat (2026-09, reporting). aider edit-format leaderboard; diff-format success by model
- [aider edit-format taxonomy (whole/diff/udiff/patch/search-replace)](https://aider.chat/docs/more/edit-formats.html) — aider.chat (2026-09, primary). aider edit-format taxonomy (whole/diff/udiff/patch/search-replace)
- [aider issue #625 — the two-llms debate](https://github.com/Aider-AI/aider/issues/625) — github (2024-05-30, repo). debate of cursor-style apply; gauthier "2 llms" latency/failure objections (june 4 comment)
- [upgraded claude 3.5 sonnet + computer use](https://www.anthropic.com/news/3-5-models-and-computer-use) — anthropic (2024-10-22, reporting). upgraded claude 3.5 sonnet; text_editor_20241022 tool
- [swe-bench + str_replace_editor details](https://www.anthropic.com/news/swe-bench-sonnet) — anthropic (2024-10-30, reporting). str_replace_editor details; "highest reliability with string replacement"
- [the infra pod on apple podcasts](https://podcasts.apple.com/in/podcast/coding-agents-need-infra-to-apply-code-changes-chat/id1689151910?i=1000748941076) — apple podcasts (2026-02-09, interview). infra pod episode w/ tejas; show notes incl. "why cursor removed fast apply"
- [aws morph case study](https://aws.amazon.com/solutions/case-studies/morph-case-study/) — aws.amazon.com (2025-12, reporting). aws: morph custom inference engine on nvidia h100; 1,000→10,000 tok/s; middleware framing
- [cognition × cerebras case study](https://www.cerebras.ai/blog/case-study-cognition-x-cerebras) — cerebras.ai (2025, reporting). cognition swe-grep deployment on cerebras; fast-inference context
- [don't build multi-agents](https://cognition.ai/blog/dont-build-multi-agents) — cognition.ai (2025-06-12, reporting). "today, the edit decision-making and applying are more often done by a single model in one action."
- [swe-grep and swe-grep-mini](https://cognition.com/blog/swe-grep) — cognition.com (2025-10-16, reporting). swe-grep/swe-grep-mini rl retrieval subagents on cerebras
- [composio podcast notes on tejas](https://www.linkedin.com/posts/composiohq_tejas-bhakta-founder-of-morph-joins-julia-activity-7447015735895146496-ME9n) — linkedin (2025-12, interview). tejas interview notes: two-person team, gpu bill 8× salary, speed→conversion claim
- [continue apply-role docs](https://docs.continue.dev/customize/model-roles/apply) — docs.continue.dev (2026-09, primary). continue apply role recommends morph fast apply or relace instant apply
- [mercury coder in continue](https://www.continue.dev/inception/mercury-coder) — continue.dev (2026-09, primary). mercury coder integration notes (speed-positioned code model)
- [tejas-authored apply-role docs incl. morph](https://github.com/continuedev/continue/pull/4663) — github (2025-03-15, repo). tejas-authored apply-role docs incl. morph; merged 2025-04-29
- [editing files at 1000 tokens per second](https://cursor.com/blog/instant-apply) — cursor (2024-05-14, reporting). original fast-apply post: ~1,000 tok/s, speculative edits, fine-tuned 70b
- [instant apply, wayback copy](https://web.archive.org/web/20240823050616/https:/www.cursor.com/blog/instant-apply) — wayback machine (2024-08-23, archive). archived copy preserving original text
- [databutton testimonial: 82%→96%](https://www.linkedin.com/company/databutton) — linkedin (2025-09-16, primary). databutton testimonial: apply accuracy 82%→96% (reported via linkedin)
- [build better agents with morphllm](https://dev.to/flyio/build-better-agents-with-morphllm-4mc8) — dev.to (2025-08-26, reporting). fly.io morph integration walkthrough
- [engineering.fyi mirror of the fly.io tutorial](https://www.engineering.fyi/article/build-better-agents-with-morphllm) — engineering.fyi (2025, reporting). mirror of fly.io morph tutorial
- [morphllm launches (yc launch repost)](https://www.fondo.com/blog/morphllm-launches) — fondo.com (2026-02-09, reporting). republished yc launch copy incl. fortune-50 staff-eng quote and self-host link
- [start pod notes](https://fondo.com/blog/start-pod-tejas-bhakta) — fondo.com (2026, interview). start pod: "agents don't need bigger models. they need better tools."
- [why ai tools rewrite full files instead of diffs](https://anishgandhi.com/why-ai-tools-dont-use-diffs/) — anishgandhi.com (2026, reference). why tools prefer full rewrites; ~400-line crossover; converge of aider/cursor/morph
- [how ai coding tools edit code under the hood](https://anishgandhi.com/how-ai-coding-tools-edit-code/) — anishgandhi.com (2026, reference). two-model edit anatomy; the "seam" where models disagree
- [ast edits benchmark](https://geometricagi.github.io/2026/04/02/ast-edits.html) — geometricagi.github.io (2026-04-02, reporting). edit-format benchmark incl. whole-file/search-replace/patch/ast
- [grok cli morph integration](https://www.npmjs.com/package/@phuetz/grok-cli) — npm (2025, community). community cli integrating morph apply (reported)
- [the hashline harness benchmark](https://mywrittenword.com/2026/04/12/hashline-edit-tool-harness-problem-llm-coding-benchmark-2026/) — mywrittenword.com (2026-04-12, reference). can bölük hashline harness benchmark; grok code fast 1 6.7%→68.3%
- [hn algolia comment corpus for morph](https://hn.algolia.com/?query=morph%20apply&type=comment) — hn.algolia.com (2026-09, community). algolia hn comment corpus for morph mentions incl. 2025-09-15 10k-tok/s self-post
- [hn submission of cursor 1000-tok/s post (10 pts)](https://news.ycombinator.com/item?id=40580258) — hacker news (2024-06-05, community). hn submission of cursor 1000-tok/s post (10 pts)
- [cursor removed its instant apply post](https://news.ycombinator.com/item?id=41687334) — hacker news (2024-09-29, community). "cursorai just removed technical blog post about their instant apply algorithm" (2 pts)
- [codegen research on hn](https://news.ycombinator.com/item?id=48633230) — hacker news (2026-06-22, community). "optimizing models to be fast at codegen" submission (2 pts)
- [everything is model[s] on hn](https://news.ycombinator.com/item?id=45488424) — hacker news (2025-10-06, community). "everything is model[s]" submission (1 pt)
- [relace a year of fast apply thread (48 pts, 6 comments)](https://news.ycombinator.com/item?id=45749763) — hacker news (2025-10-29, community). relace "a year of fast apply" thread (48 pts, 6 comments)
- [relace compaction post submission (6 pts)](https://news.ycombinator.com/item?id=48895729) — hacker news (2026-07-13, community). relace compaction post submission (6 pts)
- [launch hn: relace (111 pts, 51 comments)](https://news.ycombinator.com/item?id=44108206) — hacker news (2025-05-27, community). launch hn: relace (111 pts, 51 comments)
- [relace were trying to make kimi k3 cheaper (3 pts)](https://news.ycombinator.com/item?id=49377468) — hacker news (2026-08-20, community). relace "we're trying to make kimi k3 cheaper" (3 pts)
- [swe-grep hn thread (97 pts, 31 comments)](https://news.ycombinator.com/item?id=45607822) — hacker news (2025-10-16, community). swe-grep hn thread (97 pts, 31 comments)
- [mercury api availability](https://www.inceptionlabs.ai/blog/introducing-inception-api) — inceptionlabs.ai (2025, reporting). mercury api availability; speed positioning
- [coding agents need infra to apply code changes](https://podscripts.co/podcasts/the-infra-pod/coding-agents-need-infra-to-apply-code-changes-chat-with-tejas-from-morph) — podscripts.co (2026-02-09, interview). tejas interview: pivots, feb-2025 start, 70–80% prior, 30–40% lazy-prompt gain, b2b sdk/mcp, ~4× cerebras claim
- [kilo code release notes](https://blog.kilo.ai/p/terminal-command-generation-virtual) — blog.kilo.ai (2025-08, reporting). kilo release notes incl. v4.71.0 experimental morph fast apply (contributor tejas)
- [open-source fastapply pipeline + dataset powering softgen](https://github.com/kortix-ai/fast-apply) — github (2024-10, repo). open-source fastapply pipeline + dataset powering softgen
- [fastapply-1.5b model card](https://huggingface.co/Kortix/FastApply-1.5B-v1.0) — huggingface.co (2024-10, reference). fastapply-1.5b (qwen2.5-coder finetune) weights
- [mercury diffusion code models](https://arxiv.org/html/2506.17298v1) — arxiv.org (2025-06, reporting). mercury diffusion code models; ~1,109/737 tok/s mini/small on h100
- [apply endpoint spec](https://docs.morphllm.com/api-reference/endpoint/apply) — docs.morphllm.com (2026-09, primary). apply endpoint spec; openai-compatible surface
- [how morph scaled to 10,000 tok/s on aws](https://www.morphllm.com/blog/morph-aws-case-study) — morphllm.com (2025-12, reporting). morph's aws writeup: <400 ms 15k-token refactors; binance 50–70% gains (vendor-reported)
- [flash compact launch](https://www.morphllm.com/blog/compact-sdk) — morphllm.com (2026, reporting). flash compact launch: 33,000 tok/s verbatim compaction
- [compact sdk docs](https://docs.morphllm.com/sdk/components/compact) — docs.morphllm.com (2026-09, primary). morph-compactor: 33,000 tok/s, 1m ctx, 50–70% reduction
- [compact product page](https://www.morphllm.com/products/compact) — morphllm.com (2026-09, primary). compact marketing page (spec inconsistencies vs. docs noted)
- [one backbone, many reflexes](https://www.morphllm.com/blog/reflex-inference-engine) — morphllm.com (2026-06-23, reporting). "one backbone, many reflexes" multi-head classifier engine
- [fast apply sdk docs](https://docs.morphllm.com/sdk/components/fast-apply) — docs.morphllm.com (2026-09, primary). sdk component docs (spec variants vs. other pages)
- [warpgrep benchmark page](https://www.morphllm.com/benchmarks/warp-grep) — morphllm.com (2026-09, reporting). vendor warpgrep benchmark page
- [gpt-4.1 launch](https://openai.com/index/gpt-4-1/) — openai (2025-04-14, reporting). gpt-4.1 launch; major gains claimed on aider polyglot diff benchmark
- [morph models on openrouter](https://openrouter.ai/morph/morph-v3-fast) — openrouter.ai (2026-09, reference). morph models on openrouter; pricing + observed-throughput context (sibling: /morph/morph-v3-large)
- [osmosis-apply-1.7b model card](https://huggingface.co/osmosis-ai/Osmosis-Apply-1.7B) — huggingface.co (2025, reference). open-source apply model (qwen3-1.7b base)
- [mcp server for osmosis-apply](https://github.com/osmosis-ai/osmosis-apply-1.7b-mcp) — github (2025, repo). mcp server for osmosis-apply
- [rl for code-merging writeup](https://osmosis.ai/blogs/applying-rl-improving-code-merging) — osmosis.ai (2025, reporting). rl for code-merging writeup
- [fast apply models are already dead](https://pashpashpash.substack.com/p/fast-apply-models-are-already-dead) — pashpashpash.substack.com (2025-06-20, reporting). bear-case post; gauthier quote; tejas "lol, i never said that" reply (sep 10, 2025)
- [hash-anchored edit tools](https://github.com/lkraider/pi-linehash-edit/) — github (2026, repo). hash-anchored edit tools; cites bölük 16-model harness benchmark
- [relace series a announcement](https://relace.ai/blog/series-a-announcement) — relace.ai (2025-10-08, reporting). $23m series a led by a16z; matrix + yc participating
- [mercury coder small on vercel ai gateway](https://vercel.com/ai-gateway/models/mercury-coder-small) — vercel (2026-09, reporting). mercury coder small on vercel ai gateway
- [relace launch yc: cheaper/faster/more-reliable codegen models](https://www.ycombinator.com/launches/NV5-relace-cheaper-faster-and-more-reliable-ai-code-generation) — ycombinator.com (2025-05, primary). relace launch yc: cheaper/faster/more-reliable codegen models
- [relace yc profile](https://www.ycombinator.com/companies/relace) — ycombinator.com (2026-09, primary). yc w23 profile
- [about page footer — © 2025 autoinfra, inc.](https://arquivo.pt/wayback/20251028055508/https://www.morphllm.com/about) — arquivo.pt (2025-10-28, archive). footer: "© 2025 autoinfra, inc."
- [blog index snapshot with post dates](https://arquivo.pt/wayback/20251011053746/https://www.morphllm.com/blog) — arquivo.pt (2025-10-11, archive). blog index snapshot with post dates
- [homepage snapshot: benchmark tables, hundreds of startups and enterprises](https://arquivo.pt/wayback/20251011053800/https://www.morphllm.com/) — arquivo.pt (2025-10-11, archive). homepage snapshot: benchmark tables, "hundreds of startups and enterprises"
- [tejas bhakta on beamstart](https://beamstart.com/@tejas-bhakta-17291825058113) — beamstart.com (undated, reference). tesla autopilot/energy 3yr bio; motel/$48k website; ucsd
- [fast apply benchmark page](https://www.morphllm.com/benchmarks/fast-apply) — morphllm.com (2026, primary). self-reported benchmark table; llm-judged methodology
- [we hit 10,500 tokens/sec on b200](https://www.morphllm.com/blog/morph-breaks-10k-barrier) — morphllm.com (2025-09-15, primary). 10,500 tok/s on b200; custom kernels + speculative execution
- [best practices for building coding agents with morph](https://www.morphllm.com/blog/best-practices) — morphllm.com (2025-05-02, primary). agent-builder best practices
- [browser verification](https://www.morphllm.com/blog/browser-verification) — morphllm.com (2026-01-23, primary). glance: rl browser-testing agent
- [optimizing models to be fast at codegen](https://www.morphllm.com/blog/codegen-inference-research) — morphllm.com (2026-06, primary). speculator economics; 1.93x vs 100x; 33k compaction; serves qwen/glm/deepseek/minimax
- [morph ❤️ continue](https://www.morphllm.com/blog/morph-continue) — morphllm.com (2025-05-28, primary). continue integration announcement
- [how cursor composer and apply work](https://www.morphllm.com/blog/cursor-composer-and-apply) — morphllm.com (2024-02-15, primary). analysis of cursor apply internals
- [diffs vs fast apply](https://www.morphllm.com/blog/diffs-vs-fast-apply) — morphllm.com (2025-05-28, primary). fast apply vs diffs argument
- [everything is models](https://www.morphllm.com/blog/everything-is-models) — morphllm.com (2025-10-10, primary). "everything is models" thesis
- [faster agents](https://www.morphllm.com/blog/faster-agents) — morphllm.com (2024-02-13, primary). earliest dated post (predates domain — migrated/backdated)
- [morph gets faster](https://www.morphllm.com/blog/morph-gets-faster) — morphllm.com (2025-06-15, primary). speed post; title later showed 10.5k (update ambiguity)
- [llms are bad at being forced](https://www.morphllm.com/blog/llms-bad-at-being-forced) — morphllm.com (2025-10-10, primary). structured-output critique
- [self hosting](https://www.morphllm.com/blog/self-hosting) — morphllm.com (2025-05-11, primary). self-host offering
- [warpgrep v2](https://www.morphllm.com/blog/warpgrep-v2) — morphllm.com (2026-03-31, primary). warpgrep v2 launch; swe-bench pro claims
- [what is morph fast apply?](https://www.morphllm.com/blog/what-is-morph-for) — morphllm.com (2025-08-22, primary). explainer post


Published by @ben in profile revision 1.
Original document: https://hraness.com/morph/sources
Original credited author: Hraness
AI-drafted at Ben Guo's direct request and credited to Hraness; every claim links to its cataloged source.

## Structured source and record data

````json
{"anchors":[{"id":"catalog","originalUrl":"https://hraness.com/morph/sources#catalog","title":"the catalog"}],"provenance":{"draftingDisclosure":"AI-drafted at Ben Guo's direct request and credited to Hraness; every claim links to its cataloged source.","originalAuthor":"Hraness","originalRevision":"03eb7fbb7a99f893f57f874434e2dc368ebc3c97b5ac46505de3895f363913af","originalUrl":"https://hraness.com/morph/sources","source":"hraness"},"records":[],"sources":[{"author":null,"id":"source-d458eeb910e0d82ec91d","note":"aider edit-format leaderboard; diff-format success by model","originalId":"src-ci-aider-edit-bench","published":{"precision":"month","text":"2026-09"},"publisher":"aider.chat","title":"aider edit-format leaderboard","type":"reporting","url":"https://aider.chat/docs/leaderboards/edit.html"},{"author":null,"id":"source-f6b1b06f38f12d48dd87","note":"aider edit-format taxonomy (whole/diff/udiff/patch/search-replace)","originalId":"src-ci-aider-edit-formats","published":{"precision":"month","text":"2026-09"},"publisher":"aider.chat","title":"aider edit-format taxonomy (whole/diff/udiff/patch/search-replace)","type":"primary","url":"https://aider.chat/docs/more/edit-formats.html"},{"author":null,"id":"source-847d0cb8d8642ee26fbd","note":"debate of cursor-style apply; gauthier \"2 llms\" latency/failure objections (june 4 comment)","originalId":"src-ci-aider-issue-625","published":{"precision":"day","text":"2024-05-30"},"publisher":"github","title":"aider issue #625 — the two-llms debate","type":"repo","url":"https://github.com/Aider-AI/aider/issues/625"},{"author":null,"id":"source-a6a2f3d0c41884a30edb","note":"upgraded claude 3.5 sonnet; text_editor_20241022 tool","originalId":"src-ci-anthropic-35","published":{"precision":"day","text":"2024-10-22"},"publisher":"anthropic","title":"upgraded claude 3.5 sonnet + computer use","type":"reporting","url":"https://www.anthropic.com/news/3-5-models-and-computer-use"},{"author":null,"id":"source-e0bfbd0194447b48d569","note":"str_replace_editor details; \"highest reliability with string replacement\"","originalId":"src-ci-anthropic-swe-bench","published":{"precision":"day","text":"2024-10-30"},"publisher":"anthropic","title":"swe-bench + str_replace_editor details","type":"reporting","url":"https://www.anthropic.com/news/swe-bench-sonnet"},{"author":null,"id":"source-ebf2b48be88d4ec95e48","note":"infra pod episode w/ tejas; show notes incl. \"why cursor removed fast apply\"","originalId":"src-ci-apple-infra-pod","published":{"precision":"day","text":"2026-02-09"},"publisher":"apple podcasts","title":"the infra pod on apple podcasts","type":"interview","url":"https://podcasts.apple.com/in/podcast/coding-agents-need-infra-to-apply-code-changes-chat/id1689151910?i=1000748941076"},{"author":null,"id":"source-1f9811793a2cd993a44e","note":"aws: morph custom inference engine on nvidia h100; 1,000→10,000 tok/s; middleware framing","originalId":"src-ci-aws-morph","published":{"precision":"month","text":"2025-12"},"publisher":"aws.amazon.com","title":"aws morph case study","type":"reporting","url":"https://aws.amazon.com/solutions/case-studies/morph-case-study/"},{"author":null,"id":"source-fd0700759758bdf9bc93","note":"cognition swe-grep deployment on cerebras; fast-inference context","originalId":"src-ci-cerebras-cognition","published":{"precision":"year","text":"2025"},"publisher":"cerebras.ai","title":"cognition × cerebras case study","type":"reporting","url":"https://www.cerebras.ai/blog/case-study-cognition-x-cerebras"},{"author":null,"id":"source-8ceccee1854c74b67eab","note":"\"today, the edit decision-making and applying are more often done by a single model in one action.\"","originalId":"src-ci-cognition-multi-agents","published":{"precision":"day","text":"2025-06-12"},"publisher":"cognition.ai","title":"don't build multi-agents","type":"reporting","url":"https://cognition.ai/blog/dont-build-multi-agents"},{"author":null,"id":"source-a81bb73f022808842172","note":"swe-grep/swe-grep-mini rl retrieval subagents on cerebras","originalId":"src-ci-cognition-swe-grep","published":{"precision":"day","text":"2025-10-16"},"publisher":"cognition.com","title":"swe-grep and swe-grep-mini","type":"reporting","url":"https://cognition.com/blog/swe-grep"},{"author":null,"id":"source-673fdcf2ba99bbd22062","note":"tejas interview notes: two-person team, gpu bill 8× salary, speed→conversion claim","originalId":"src-ci-composio-podcast","published":{"precision":"month","text":"2025-12"},"publisher":"linkedin","title":"composio podcast notes on tejas","type":"interview","url":"https://www.linkedin.com/posts/composiohq_tejas-bhakta-founder-of-morph-joins-julia-activity-7447015735895146496-ME9n"},{"author":null,"id":"source-ef5fd87c63999b039c9b","note":"continue apply role recommends morph fast apply or relace instant apply","originalId":"src-ci-continue-apply-docs","published":{"precision":"month","text":"2026-09"},"publisher":"docs.continue.dev","title":"continue apply-role docs","type":"primary","url":"https://docs.continue.dev/customize/model-roles/apply"},{"author":null,"id":"source-f489fa21f228916bece1","note":"mercury coder integration notes (speed-positioned code model)","originalId":"src-ci-continue-mercury","published":{"precision":"month","text":"2026-09"},"publisher":"continue.dev","title":"mercury coder in continue","type":"primary","url":"https://www.continue.dev/inception/mercury-coder"},{"author":null,"id":"source-692cd9123535a762ead4","note":"tejas-authored apply-role docs incl. morph; merged 2025-04-29","originalId":"src-ci-continue-pr-4663","published":{"precision":"day","text":"2025-03-15"},"publisher":"github","title":"tejas-authored apply-role docs incl. morph","type":"repo","url":"https://github.com/continuedev/continue/pull/4663"},{"author":null,"id":"source-266f972638a904162eba","note":"original fast-apply post: ~1,000 tok/s, speculative edits, fine-tuned 70b","originalId":"src-ci-cursor-instant-apply","published":{"precision":"day","text":"2024-05-14"},"publisher":"cursor","title":"editing files at 1000 tokens per second","type":"reporting","url":"https://cursor.com/blog/instant-apply"},{"author":null,"id":"source-26734fe44cf8d928b532","note":"archived copy preserving original text","originalId":"src-ci-cursor-instant-apply-wayback","published":{"precision":"day","text":"2024-08-23"},"publisher":"wayback machine","title":"instant apply, wayback copy","type":"archive","url":"https://web.archive.org/web/20240823050616/https:/www.cursor.com/blog/instant-apply"},{"author":null,"id":"source-2aec268f4a2ae19e7829","note":"databutton testimonial: apply accuracy 82%→96% (reported via linkedin)","originalId":"src-ci-databutton-li","published":{"precision":"day","text":"2025-09-16"},"publisher":"linkedin","title":"databutton testimonial: 82%→96%","type":"primary","url":"https://www.linkedin.com/company/databutton"},{"author":null,"id":"source-21f8667c4fee67f0619b","note":"fly.io morph integration walkthrough","originalId":"src-ci-devto-flyio","published":{"precision":"day","text":"2025-08-26"},"publisher":"dev.to","title":"build better agents with morphllm","type":"reporting","url":"https://dev.to/flyio/build-better-agents-with-morphllm-4mc8"},{"author":null,"id":"source-0bbae248264bb4da0ed0","note":"mirror of fly.io morph tutorial","originalId":"src-ci-eng-fyi","published":{"precision":"year","text":"2025"},"publisher":"engineering.fyi","title":"engineering.fyi mirror of the fly.io tutorial","type":"reporting","url":"https://www.engineering.fyi/article/build-better-agents-with-morphllm"},{"author":null,"id":"source-5e52d28d6ca21342da5a","note":"republished yc launch copy incl. fortune-50 staff-eng quote and self-host link","originalId":"src-ci-fondo-morph","published":{"precision":"day","text":"2026-02-09"},"publisher":"fondo.com","title":"morphllm launches (yc launch repost)","type":"reporting","url":"https://www.fondo.com/blog/morphllm-launches"},{"author":null,"id":"source-a5773f38ef217edd4552","note":"start pod: \"agents don't need bigger models. they need better tools.\"","originalId":"src-ci-fondo-start-pod","published":{"precision":"year","text":"2026"},"publisher":"fondo.com","title":"start pod notes","type":"interview","url":"https://fondo.com/blog/start-pod-tejas-bhakta"},{"author":null,"id":"source-5cfd2e54566a8146a7dc","note":"why tools prefer full rewrites; ~400-line crossover; converge of aider/cursor/morph","originalId":"src-ci-gandhi-diffs","published":{"precision":"year","text":"2026"},"publisher":"anishgandhi.com","title":"why ai tools rewrite full files instead of diffs","type":"reference","url":"https://anishgandhi.com/why-ai-tools-dont-use-diffs/"},{"author":null,"id":"source-760bf9dec69ca8fe3f7f","note":"two-model edit anatomy; the \"seam\" where models disagree","originalId":"src-ci-gandhi-editing","published":{"precision":"year","text":"2026"},"publisher":"anishgandhi.com","title":"how ai coding tools edit code under the hood","type":"reference","url":"https://anishgandhi.com/how-ai-coding-tools-edit-code/"},{"author":null,"id":"source-4574a2d766ac5d7820cd","note":"edit-format benchmark incl. whole-file/search-replace/patch/ast","originalId":"src-ci-geometric-ast","published":{"precision":"day","text":"2026-04-02"},"publisher":"geometricagi.github.io","title":"ast edits benchmark","type":"reporting","url":"https://geometricagi.github.io/2026/04/02/ast-edits.html"},{"author":null,"id":"source-af58fc55d1624e34d0e4","note":"community cli integrating morph apply (reported)","originalId":"src-ci-grok-cli","published":{"precision":"year","text":"2025"},"publisher":"npm","title":"grok cli morph integration","type":"community","url":"https://www.npmjs.com/package/@phuetz/grok-cli"},{"author":null,"id":"source-b8eaadc3ae300b5cfef3","note":"can bölük hashline harness benchmark; grok code fast 1 6.7%→68.3%","originalId":"src-ci-hashline-mww","published":{"precision":"day","text":"2026-04-12"},"publisher":"mywrittenword.com","title":"the hashline harness benchmark","type":"reference","url":"https://mywrittenword.com/2026/04/12/hashline-edit-tool-harness-problem-llm-coding-benchmark-2026/"},{"author":null,"id":"source-8e8eb4bc44886e6937c2","note":"algolia hn comment corpus for morph mentions incl. 2025-09-15 10k-tok/s self-post","originalId":"src-ci-hn-algolia-comments","published":{"precision":"month","text":"2026-09"},"publisher":"hn.algolia.com","title":"hn algolia comment corpus for morph","type":"community","url":"https://hn.algolia.com/?query=morph%20apply&type=comment"},{"author":null,"id":"source-319490e4806744ba0963","note":"hn submission of cursor 1000-tok/s post (10 pts)","originalId":"src-ci-hn-cursor-1000","published":{"precision":"day","text":"2024-06-05"},"publisher":"hacker news","title":"hn submission of cursor 1000-tok/s post (10 pts)","type":"community","url":"https://news.ycombinator.com/item?id=40580258"},{"author":null,"id":"source-6e6340d3066252a86510","note":"\"cursorai just removed technical blog post about their instant apply algorithm\" (2 pts)","originalId":"src-ci-hn-cursor-removed","published":{"precision":"day","text":"2024-09-29"},"publisher":"hacker news","title":"cursor removed its instant apply post","type":"community","url":"https://news.ycombinator.com/item?id=41687334"},{"author":null,"id":"source-eb0e680e1e1f0c7862e5","note":"\"optimizing models to be fast at codegen\" submission (2 pts)","originalId":"src-ci-hn-morph-codegen","published":{"precision":"day","text":"2026-06-22"},"publisher":"hacker news","title":"codegen research on hn","type":"community","url":"https://news.ycombinator.com/item?id=48633230"},{"author":null,"id":"source-8a1f364b353c8d9f6d46","note":"\"everything is model[s]\" submission (1 pt)","originalId":"src-ci-hn-morph-eim","published":{"precision":"day","text":"2025-10-06"},"publisher":"hacker news","title":"everything is model[s] on hn","type":"community","url":"https://news.ycombinator.com/item?id=45488424"},{"author":null,"id":"source-15be7653ac44651e701a","note":"relace \"a year of fast apply\" thread (48 pts, 6 comments)","originalId":"src-ci-hn-relace-apply3","published":{"precision":"day","text":"2025-10-29"},"publisher":"hacker news","title":"relace a year of fast apply thread (48 pts, 6 comments)","type":"community","url":"https://news.ycombinator.com/item?id=45749763"},{"author":null,"id":"source-ed5bd51aab8847693074","note":"relace compaction post submission (6 pts)","originalId":"src-ci-hn-relace-compact","published":{"precision":"day","text":"2026-07-13"},"publisher":"hacker news","title":"relace compaction post submission (6 pts)","type":"community","url":"https://news.ycombinator.com/item?id=48895729"},{"author":null,"id":"source-7806531811af8f84b9b1","note":"launch hn: relace (111 pts, 51 comments)","originalId":"src-ci-hn-relace-launch","published":{"precision":"day","text":"2025-05-27"},"publisher":"hacker news","title":"launch hn: relace (111 pts, 51 comments)","type":"community","url":"https://news.ycombinator.com/item?id=44108206"},{"author":null,"id":"source-9396d8f9d061adbd8fef","note":"relace \"we're trying to make kimi k3 cheaper\" (3 pts)","originalId":"src-ci-hn-relace-packs","published":{"precision":"day","text":"2026-08-20"},"publisher":"hacker news","title":"relace were trying to make kimi k3 cheaper (3 pts)","type":"community","url":"https://news.ycombinator.com/item?id=49377468"},{"author":null,"id":"source-a7380b2d95b0a22a74ae","note":"swe-grep hn thread (97 pts, 31 comments)","originalId":"src-ci-hn-swe-grep","published":{"precision":"day","text":"2025-10-16"},"publisher":"hacker news","title":"swe-grep hn thread (97 pts, 31 comments)","type":"community","url":"https://news.ycombinator.com/item?id=45607822"},{"author":null,"id":"source-e065b5f9a00dd95105ea","note":"mercury api availability; speed positioning","originalId":"src-ci-inception-api","published":{"precision":"year","text":"2025"},"publisher":"inceptionlabs.ai","title":"mercury api availability","type":"reporting","url":"https://www.inceptionlabs.ai/blog/introducing-inception-api"},{"author":null,"id":"source-cd996b7c516a0f39611d","note":"tejas interview: pivots, feb-2025 start, 70–80% prior, 30–40% lazy-prompt gain, b2b sdk/mcp, ~4× cerebras claim","originalId":"src-ci-infra-pod","published":{"precision":"day","text":"2026-02-09"},"publisher":"podscripts.co","title":"coding agents need infra to apply code changes","type":"interview","url":"https://podscripts.co/podcasts/the-infra-pod/coding-agents-need-infra-to-apply-code-changes-chat-with-tejas-from-morph"},{"author":null,"id":"source-5a405bfe68da7835eb83","note":"kilo release notes incl. v4.71.0 experimental morph fast apply (contributor tejas)","originalId":"src-ci-kilo-blog","published":{"precision":"month","text":"2025-08"},"publisher":"blog.kilo.ai","title":"kilo code release notes","type":"reporting","url":"https://blog.kilo.ai/p/terminal-command-generation-virtual"},{"author":null,"id":"source-7adf2a31adebf2159329","note":"open-source fastapply pipeline + dataset powering softgen","originalId":"src-ci-kortix-gh","published":{"precision":"month","text":"2024-10"},"publisher":"github","title":"open-source fastapply pipeline + dataset powering softgen","type":"repo","url":"https://github.com/kortix-ai/fast-apply"},{"author":null,"id":"source-46a0870d320e42d3c29d","note":"fastapply-1.5b (qwen2.5-coder finetune) weights","originalId":"src-ci-kortix-hf","published":{"precision":"month","text":"2024-10"},"publisher":"huggingface.co","title":"fastapply-1.5b model card","type":"reference","url":"https://huggingface.co/Kortix/FastApply-1.5B-v1.0"},{"author":null,"id":"source-1356f0570e8368d898e4","note":"mercury diffusion code models; ~1,109/737 tok/s mini/small on h100","originalId":"src-ci-mercury-arxiv","published":{"precision":"month","text":"2025-06"},"publisher":"arxiv.org","title":"mercury diffusion code models","type":"reporting","url":"https://arxiv.org/html/2506.17298v1"},{"author":null,"id":"source-ac7171b8b4952be265aa","note":"apply endpoint spec; openai-compatible surface","originalId":"src-ci-morph-api-apply","published":{"precision":"month","text":"2026-09"},"publisher":"docs.morphllm.com","title":"apply endpoint spec","type":"primary","url":"https://docs.morphllm.com/api-reference/endpoint/apply"},{"author":null,"id":"source-98f716504839ae5f64fe","note":"morph's aws writeup: <400 ms 15k-token refactors; binance 50–70% gains (vendor-reported)","originalId":"src-ci-morph-aws-case","published":{"precision":"month","text":"2025-12"},"publisher":"morphllm.com","title":"how morph scaled to 10,000 tok/s on aws","type":"reporting","url":"https://www.morphllm.com/blog/morph-aws-case-study"},{"author":null,"id":"source-e784d030cd488dfe11d3","note":"flash compact launch: 33,000 tok/s verbatim compaction","originalId":"src-ci-morph-compact-blog","published":{"precision":"year","text":"2026"},"publisher":"morphllm.com","title":"flash compact launch","type":"reporting","url":"https://www.morphllm.com/blog/compact-sdk"},{"author":null,"id":"source-a92ffb77a1f3714b8bb2","note":"morph-compactor: 33,000 tok/s, 1m ctx, 50–70% reduction","originalId":"src-ci-morph-compact-docs","published":{"precision":"month","text":"2026-09"},"publisher":"docs.morphllm.com","title":"compact sdk docs","type":"primary","url":"https://docs.morphllm.com/sdk/components/compact"},{"author":null,"id":"source-962a5e011f349884e332","note":"compact marketing page (spec inconsistencies vs. docs noted)","originalId":"src-ci-morph-compact-product","published":{"precision":"month","text":"2026-09"},"publisher":"morphllm.com","title":"compact product page","type":"primary","url":"https://www.morphllm.com/products/compact"},{"author":null,"id":"source-1825defbea8ea58ab7bb","note":"\"one backbone, many reflexes\" multi-head classifier engine","originalId":"src-ci-morph-reflex-blog","published":{"precision":"day","text":"2026-06-23"},"publisher":"morphllm.com","title":"one backbone, many reflexes","type":"reporting","url":"https://www.morphllm.com/blog/reflex-inference-engine"},{"author":null,"id":"source-eeef06e6cab3ae22c6f4","note":"sdk component docs (spec variants vs. other pages)","originalId":"src-ci-morph-sdk-fast-apply","published":{"precision":"month","text":"2026-09"},"publisher":"docs.morphllm.com","title":"fast apply sdk docs","type":"primary","url":"https://docs.morphllm.com/sdk/components/fast-apply"},{"author":null,"id":"source-20b1190233d3c21420ae","note":"vendor warpgrep benchmark page","originalId":"src-ci-morph-warpgrep-bench","published":{"precision":"month","text":"2026-09"},"publisher":"morphllm.com","title":"warpgrep benchmark page","type":"reporting","url":"https://www.morphllm.com/benchmarks/warp-grep"},{"author":null,"id":"source-1d333920bffd083c28e9","note":"gpt-4.1 launch; major gains claimed on aider polyglot diff benchmark","originalId":"src-ci-openai-gpt41","published":{"precision":"day","text":"2025-04-14"},"publisher":"openai","title":"gpt-4.1 launch","type":"reporting","url":"https://openai.com/index/gpt-4-1/"},{"author":null,"id":"source-2e94b35ed218a876aac1","note":"morph models on openrouter; pricing + observed-throughput context (sibling: /morph/morph-v3-large)","originalId":"src-ci-openrouter-morph","published":{"precision":"month","text":"2026-09"},"publisher":"openrouter.ai","title":"morph models on openrouter","type":"reference","url":"https://openrouter.ai/morph/morph-v3-fast"},{"author":null,"id":"source-769f14e24f769b4a0cde","note":"open-source apply model (qwen3-1.7b base)","originalId":"src-ci-osmosis-hf","published":{"precision":"year","text":"2025"},"publisher":"huggingface.co","title":"osmosis-apply-1.7b model card","type":"reference","url":"https://huggingface.co/osmosis-ai/Osmosis-Apply-1.7B"},{"author":null,"id":"source-ff84f4b5c93b3becaa65","note":"mcp server for osmosis-apply","originalId":"src-ci-osmosis-mcp","published":{"precision":"year","text":"2025"},"publisher":"github","title":"mcp server for osmosis-apply","type":"repo","url":"https://github.com/osmosis-ai/osmosis-apply-1.7b-mcp"},{"author":null,"id":"source-04533179f5e9605ab1d7","note":"rl for code-merging writeup","originalId":"src-ci-osmosis-rl","published":{"precision":"year","text":"2025"},"publisher":"osmosis.ai","title":"rl for code-merging writeup","type":"reporting","url":"https://osmosis.ai/blogs/applying-rl-improving-code-merging"},{"author":null,"id":"source-54bfd539357adb3b87b4","note":"bear-case post; gauthier quote; tejas \"lol, i never said that\" reply (sep 10, 2025)","originalId":"src-ci-pash-dead","published":{"precision":"day","text":"2025-06-20"},"publisher":"pashpashpash.substack.com","title":"fast apply models are already dead","type":"reporting","url":"https://pashpashpash.substack.com/p/fast-apply-models-are-already-dead"},{"author":null,"id":"source-404dc449995a8dffbaa4","note":"hash-anchored edit tools; cites bölük 16-model harness benchmark","originalId":"src-ci-pi-hashline","published":{"precision":"year","text":"2026"},"publisher":"github","title":"hash-anchored edit tools","type":"repo","url":"https://github.com/lkraider/pi-linehash-edit/"},{"author":null,"id":"source-63dde4fd24be2ba1cabe","note":"$23m series a led by a16z; matrix + yc participating","originalId":"src-ci-relace-series-a","published":{"precision":"day","text":"2025-10-08"},"publisher":"relace.ai","title":"relace series a announcement","type":"reporting","url":"https://relace.ai/blog/series-a-announcement"},{"author":null,"id":"source-c67a521aaa6de65ebc3e","note":"mercury coder small on vercel ai gateway","originalId":"src-ci-vercel-mercury","published":{"precision":"month","text":"2026-09"},"publisher":"vercel","title":"mercury coder small on vercel ai gateway","type":"reporting","url":"https://vercel.com/ai-gateway/models/mercury-coder-small"},{"author":null,"id":"source-045be6bf893be31d6210","note":"relace launch yc: cheaper/faster/more-reliable codegen models","originalId":"src-ci-yc-launch-relace","published":{"precision":"month","text":"2025-05"},"publisher":"ycombinator.com","title":"relace launch yc: cheaper/faster/more-reliable codegen models","type":"primary","url":"https://www.ycombinator.com/launches/NV5-relace-cheaper-faster-and-more-reliable-ai-code-generation"},{"author":null,"id":"source-86c8eefe022000c4eb22","note":"yc w23 profile","originalId":"src-ci-yc-relace","published":{"precision":"month","text":"2026-09"},"publisher":"ycombinator.com","title":"relace yc profile","type":"primary","url":"https://www.ycombinator.com/companies/relace"},{"author":null,"id":"source-7aebfb2cf6e618b9024a","note":"footer: \"© 2025 autoinfra, inc.\"","originalId":"src-hp-arch-about-oct2025","published":{"precision":"day","text":"2025-10-28"},"publisher":"arquivo.pt","title":"about page footer — © 2025 autoinfra, inc.","type":"archive","url":"https://arquivo.pt/wayback/20251028055508/https://www.morphllm.com/about"},{"author":null,"id":"source-b62bd98c94638fd0dd0f","note":"blog index snapshot with post dates","originalId":"src-hp-arch-blog-oct2025","published":{"precision":"day","text":"2025-10-11"},"publisher":"arquivo.pt","title":"blog index snapshot with post dates","type":"archive","url":"https://arquivo.pt/wayback/20251011053746/https://www.morphllm.com/blog"},{"author":null,"id":"source-4ab26be22a8e295330c2","note":"homepage snapshot: benchmark tables, \"hundreds of startups and enterprises\"","originalId":"src-hp-arch-home-oct2025","published":{"precision":"day","text":"2025-10-11"},"publisher":"arquivo.pt","title":"homepage snapshot: benchmark tables, hundreds of startups and enterprises","type":"archive","url":"https://arquivo.pt/wayback/20251011053800/https://www.morphllm.com/"},{"author":null,"id":"source-f80ba8eddb1365ec88c5","note":"tesla autopilot/energy 3yr bio; motel/$48k website; ucsd","originalId":"src-hp-beamstart","published":{"precision":"unknown","text":"undated"},"publisher":"beamstart.com","title":"tejas bhakta on beamstart","type":"reference","url":"https://beamstart.com/@tejas-bhakta-17291825058113"},{"author":null,"id":"source-62a06a38a74e3c229045","note":"self-reported benchmark table; llm-judged methodology","originalId":"src-hp-bench-fast-apply","published":{"precision":"year","text":"2026"},"publisher":"morphllm.com","title":"fast apply benchmark page","type":"primary","url":"https://www.morphllm.com/benchmarks/fast-apply"},{"author":null,"id":"source-5f886c9538a71c5a4ab3","note":"10,500 tok/s on b200; custom kernels + speculative execution","originalId":"src-hp-blog-10k","published":{"precision":"day","text":"2025-09-15"},"publisher":"morphllm.com","title":"we hit 10,500 tokens/sec on b200","type":"primary","url":"https://www.morphllm.com/blog/morph-breaks-10k-barrier"},{"author":null,"id":"source-93fa52b10190bb0d7760","note":"agent-builder best practices","originalId":"src-hp-blog-best-practices","published":{"precision":"day","text":"2025-05-02"},"publisher":"morphllm.com","title":"best practices for building coding agents with morph","type":"primary","url":"https://www.morphllm.com/blog/best-practices"},{"author":null,"id":"source-c64d07939385812bea94","note":"glance: rl browser-testing agent","originalId":"src-hp-blog-browser-verif","published":{"precision":"day","text":"2026-01-23"},"publisher":"morphllm.com","title":"browser verification","type":"primary","url":"https://www.morphllm.com/blog/browser-verification"},{"author":null,"id":"source-81028082e2d4e3cef7aa","note":"speculator economics; 1.93x vs 100x; 33k compaction; serves qwen/glm/deepseek/minimax","originalId":"src-hp-blog-codegen-research","published":{"precision":"month","text":"2026-06"},"publisher":"morphllm.com","title":"optimizing models to be fast at codegen","type":"primary","url":"https://www.morphllm.com/blog/codegen-inference-research"},{"author":null,"id":"source-eb0c5d7e032b8f841b94","note":"continue integration announcement","originalId":"src-hp-blog-continue","published":{"precision":"day","text":"2025-05-28"},"publisher":"morphllm.com","title":"morph ❤️ continue","type":"primary","url":"https://www.morphllm.com/blog/morph-continue"},{"author":null,"id":"source-e8759e5ef898360bb236","note":"analysis of cursor apply internals","originalId":"src-hp-blog-cursor-apply","published":{"precision":"day","text":"2024-02-15"},"publisher":"morphllm.com","title":"how cursor composer and apply work","type":"primary","url":"https://www.morphllm.com/blog/cursor-composer-and-apply"},{"author":null,"id":"source-114f51059a961097df6d","note":"fast apply vs diffs argument","originalId":"src-hp-blog-diffs","published":{"precision":"day","text":"2025-05-28"},"publisher":"morphllm.com","title":"diffs vs fast apply","type":"primary","url":"https://www.morphllm.com/blog/diffs-vs-fast-apply"},{"author":null,"id":"source-b37d238fbd17881a7590","note":"\"everything is models\" thesis","originalId":"src-hp-blog-everything-models","published":{"precision":"day","text":"2025-10-10"},"publisher":"morphllm.com","title":"everything is models","type":"primary","url":"https://www.morphllm.com/blog/everything-is-models"},{"author":null,"id":"source-d092a5593a92d040ab5a","note":"earliest dated post (predates domain — migrated/backdated)","originalId":"src-hp-blog-faster-agents","published":{"precision":"day","text":"2024-02-13"},"publisher":"morphllm.com","title":"faster agents","type":"primary","url":"https://www.morphllm.com/blog/faster-agents"},{"author":null,"id":"source-586438771146fb92d17c","note":"speed post; title later showed 10.5k (update ambiguity)","originalId":"src-hp-blog-gets-faster","published":{"precision":"day","text":"2025-06-15"},"publisher":"morphllm.com","title":"morph gets faster","type":"primary","url":"https://www.morphllm.com/blog/morph-gets-faster"},{"author":null,"id":"source-50d5b4c340c570fd7460","note":"structured-output critique","originalId":"src-hp-blog-llms-forced","published":{"precision":"day","text":"2025-10-10"},"publisher":"morphllm.com","title":"llms are bad at being forced","type":"primary","url":"https://www.morphllm.com/blog/llms-bad-at-being-forced"},{"author":null,"id":"source-1aa2ab609dfc596a2b4f","note":"self-host offering","originalId":"src-hp-blog-self-hosting","published":{"precision":"day","text":"2025-05-11"},"publisher":"morphllm.com","title":"self hosting","type":"primary","url":"https://www.morphllm.com/blog/self-hosting"},{"author":null,"id":"source-00c189971ee0abfae05b","note":"warpgrep v2 launch; swe-bench pro claims","originalId":"src-hp-blog-warpgrep-v2","published":{"precision":"day","text":"2026-03-31"},"publisher":"morphllm.com","title":"warpgrep v2","type":"primary","url":"https://www.morphllm.com/blog/warpgrep-v2"},{"author":null,"id":"source-a356d7d85b600d6557d9","note":"explainer post","originalId":"src-hp-blog-what-is","published":{"precision":"day","text":"2025-08-22"},"publisher":"morphllm.com","title":"what is morph fast apply?","type":"primary","url":"https://www.morphllm.com/blog/what-is-morph-for"}]}
````