Title: fix(providers): correct pricing, deprecations, and capabilities across model catalog by waleedlatif1 · Pull Request #4990 · simstudioai/sim · GitHub
Open Graph Title: fix(providers): correct pricing, deprecations, and capabilities across model catalog by waleedlatif1 · Pull Request #4990 · simstudioai/sim
X Title: fix(providers): correct pricing, deprecations, and capabilities across model catalog by waleedlatif1 · Pull Request #4990 · simstudioai/sim
Description: Summary Two-pass validation of every static model entry in models.ts (~160 models, 12 providers + embedding/rerank pricing) against live provider docs, with secondary pricing cross-checks. Round 2 re-verified every field with a second fleet of per-provider agents, each producing a justification log covering every field, its source URL, and what was changed vs deliberately left alone (kept as a local decision log, not committed). Round 1 — pricing/deprecation sweep grok-4.20 trio price cut ($2/$6 → $1.25/$2.50), DeepSeek V4 repricing ($0.14/$0.28, 1M ctx), Gemini Deep Research output 6x fix ($2 → $12), Bedrock Mistral Large 3 4x fix ($2/$6 → $0.50/$1.50) ~30 retired/redirected models flagged deprecated (xAI May-15 batch, Gemini 2.0 shutdown, Anthropic 4.0 retirements, Mistral/Cerebras/Groq/Bedrock lifecycle) Context-window and capability corrections (sonnet-4-5/4-0 1M → 200k, invalid minimal/xhigh effort values, dead nativeStructuredOutputs on opus-4-1, renamed shut-down preview ids) Round 2 — full re-verification + remaining fixes Bedrock geo-profile bug fixed in code: getBedrockInferenceProfileId no longer prefixes the 11 models whose AWS cards say geo inference is not supported (Mistral text models, Cohere, Titan) — these previously produced invalid model IDs at runtime. Unit tests added. Bedrock pricing verified via the AWS Pricing API (marketing page is unfetchable): Nova 2 Pro/Lite, Llama 3.x/4 family, Mistral Large 2407, Ministral 3 corrections; cache-read rates added for Claude 4.x and Nova; maxOutputTokens from model cards (opus-4-1 32000 per card) azure/gpt-5.1-codex un-deprecated — round-1 stopgap was based on a wrong premise; the azure provider defaults to the Responses API, so the model works. Azure 'none' effort dropped from the 5.4 family (Azure docs enumerate 'none' support exhaustively); azure/gpt-4o deprecated (retires 2026-10-01) gpt-5.5-pro: undocumented verbosity removed and effort values aligned to documented pro-tier set (medium/high/xhigh); o3/o3-mini/o1/gpt-4.1-nano deprecated per the OpenAI deprecations page (shutdowns 2026-10-23) claude-fable-5 nativeStructuredOutputs: true — documented GA; previously fell back to prompt-injected schemas Temperature ranges corrected against API references: xAI 0–2 (was 0–1), Mistral 0–1.5 (chat-completions max per OpenAPI spec; new 1.5 slider variant in the agent block + parameterized getModelsWithTemperatureRange), Together 0–1 (their docs), Groq provider-level 0–2, deepseek-chat/Cerebras temperature exposed xAI retired slugs repriced to their redirect targets' billed rates (retirement page states redirect billing; legacy values overestimated live cost up to 6x) maxOutputTokens filled in for Groq (verified via Groq's live models API), Cerebras, Vertex (was silently falling back to 4096); recommended/speedOptimized flags normalized across providers; ollama-cloud no longer advertises unsupported tool_choice Justification logs (one per provider, maintained locally): per-model field tables with the verifying source URL and verdict, changes applied, changes deliberately not made (with reasons), and unverifiable items. Conflicting agent findings were re-verified directly against the primary source before deciding (e.g., gemini-3.1-flash-lite minimal thinking support — confirmed supported and default). Known follow-ups (documented, intentionally not in this PR) Wire reasoning_effort for xAI/Cerebras/Magistral and prompt_cache_key for Mistral, then add the corresponding capability flags Add new SKUs: deepseek-v4-flash/pro (aliases retire 2026-07-24), grok-build-0.1, mistral-medium-3-5, newer Azure/Foundry models, Gemini deep-research 04-2026 SKUs Vertex 2.5 family retires 2026-10-16 — revisit deprecation + vertex defaultModel ~Sept 2026 Cohere-on-Bedrock prices unverifiable via the Pricing API (match list prices; both models Legacy/deprecated) Type of Change Bug fix Testing All provider, block, and landing-catalog tests pass (151 + 83 in affected files; assertions updated only where they encoded the old incorrect values). New unit tests for the bedrock geo-profile logic. Typecheck and lint clean. Pre-existing providers/ test failures on staging are unrelated (verified by stash-run on clean tree). Checklist Code follows project style guidelines Self-reviewed my changes Tests added/updated and passing No new warnings introduced I confirm that I have read and agree to the terms outlined in the Contributor License Agreement (CLA)
Open Graph Description: Summary Two-pass validation of every static model entry in models.ts (~160 models, 12 providers + embedding/rerank pricing) against live provider docs, with secondary pricing cross-checks. Round 2 ...
X Description: Summary Two-pass validation of every static model entry in models.ts (~160 models, 12 providers + embedding/rerank pricing) against live provider docs, with secondary pricing cross-checks. Round 2 ...
Opengraph URL: https://github.com/simstudioai/sim/pull/4990
X: @github
Domain: github.com
| react-profiling | 0 |
| react-import-map | react,react-dom,react-dom/client,react-dom/profiling,react-is,react-reconciler,react/compiler-runtime,react/jsx-dev-runtime,react/jsx-runtime,scheduler@d5750093facf |
| route-pattern | /:user_id/:repository/pull/:id/files(.:format) |
| route-controller | pull_requests |
| route-action | files |
| fetch-nonce | v2:fc515368-7fd8-6ba8-8e2f-c3b31ca92181 |
| current-catalog-service-hash | ae870bc5e265a340912cde392f23dad3671a0a881730ffdadd82f2f57d81641b |
| request-id | A5BA:25245F:B1A5B8:F3EBDE:6AFB526F |
| html-safe-nonce | cc4a0dc3f6e36da74d195d2a99cdebe33ffc764e75a9321e0cb6c212fe98007a |
| visitor-payload | eyJyZWZlcnJlciI6IiIsInJlcXVlc3RfaWQiOiJBNUJBOjI1MjQ1RjpCMUE1Qjg6RjNFQkRFOjZBRkI1MjZGIiwidmlzaXRvcl9pZCI6IjU0NjI5OTc0Nzk0NDE3ODEzNTkiLCJyZWdpb25fZWRnZSI6ImlhZCIsInJlZ2lvbl9yZW5kZXIiOiJpYWQifQ== |
| visitor-hmac | 6dad2b0e60169947157d31b5c56a49f4f9f030dcb3e2840d9c61dd75955c93f3 |
| hovercard-subject-tag | pull_request:3851621424 |
| github-keyboard-shortcuts | repository,pull-request-list,pull-request-conversation,pull-request-files-changed,copilot |
| google-site-verification | Apib7-x98H0j5cPqHWwSMm6dNU4GmODRoqxLiDzdx9I |
| octolytics-url | https://collector.github.com/github/collect |
| analytics-location | / |
| fb:app_id | 1401488693436528 |
| apple-itunes-app | app-id=1477376905, app-argument=https://github.com/simstudioai/sim/pull/4990/files |
| twitter:image | https://avatars.githubusercontent.com/u/40672544?s=400&v=4 |
| twitter:card | summary_large_image |
| og:image | https://avatars.githubusercontent.com/u/40672544?s=400&v=4 |
| og:image:alt | Summary Two-pass validation of every static model entry in models.ts (~160 models, 12 providers + embedding/rerank pricing) against live provider docs, with secondary pricing cross-checks. Round 2 ... |
| og:site_name | GitHub |
| og:type | object |
| hostname | github.com |
| expected-hostname | github.com |
| None | 7122bb414b542dba14d71a3f0b9f6fd2abce715b9f5912d082800563e6c806d1 |
| turbo-cache-control | no-preview |
| diff-view | unified |
| go-import | github.com/simstudioai/sim git https://github.com/simstudioai/sim.git |
| octolytics-dimension-user_id | 199344406 |
| octolytics-dimension-user_login | simstudioai |
| octolytics-dimension-repository_id | 912559512 |
| octolytics-dimension-repository_nwo | simstudioai/sim |
| octolytics-dimension-repository_public | true |
| octolytics-dimension-repository_is_fork | false |
| octolytics-dimension-repository_network_root_id | 912559512 |
| octolytics-dimension-repository_network_root_nwo | simstudioai/sim |
| turbo-body-classes | logged-out env-production page-responsive full-width |
| disable-turbo | true |
| browser-stats-url | https://api.github.com/_private/browser/stats |
| browser-errors-url | https://api.github.com/_private/browser/errors |
| release | cbf66c024bb6d10e6955baa7413c54cf13a201bb |
| ui-target | full |
| theme-color | #1e2327 |
| color-scheme | light dark |
Links:
Viewport: width=device-width