06f44f509c
chore(cloudflare-ai-gateway): refresh catalog against first-party and synced sources ( #4676 )
...
* chore(cloudflare-ai-gateway): refresh catalog against first-party and synced sources
* chore(cloudflare-ai-gateway): use base_model stubs for all catalog entries
* chore(cloudflare-ai-gateway): omit experimental fast modes pending gateway billing verification
* fix(cloudflare-ai-gateway): add missing lab metadata and enforce base_model stubs
2026-08-14 00:44:12 -05:00
92d9a6d051
feat: expand benchmarks for current major models ( #3989 )
...
* feat: add Gemini 3.6 Flash and Kimi K3 benchmarks
* feat: expand current model benchmark coverage
---------
Co-authored-by: Aiden Cline <rekram1-node@users.noreply.github.com >
2026-08-03 10:04:20 -05:00
a0bb3cbda6
fix(vercel): use base_model for new catalog entries and add Thinking Machines
...
Wire Opus Fast, Realtime Whisper, and Inkling through model metadata, and
add a Thinking Machines provider for first-party Inkling access.
2026-07-16 10:22:17 -05:00
8ade7d8d24
fix(openai): split GPT-5.6 model families
2026-07-12 15:41:26 -05:00
6af0231ae6
feat(openai): add GPT-Realtime-2.1
2026-07-10 14:20:39 -05:00
084f3e1684
add sourced benchmark results for frontier models
2026-07-10 12:08:10 -05:00
6dfc39c81b
feat(openai): add GPT-5.6 models
2026-07-09 12:51:51 -05:00
101052cc58
fix(stackit): inherit canonical GPT-OSS metadata
2026-07-05 19:39:40 -05:00
5c281e4feb
feat: model descriptions
2026-07-01 14:16:49 -05:00
ac742c52e3
feat(openai): add canonical gpt-oss-120b + gpt-oss-safeguard-120b metadata
...
Provider-agnostic models/ entries for two OpenAI open-weight models that
lack them, so providers can inherit via base_model instead of full-defining.
Capability flags verified against the live Tinfoil API.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com >
2026-06-27 08:35:50 +04:00
f638d464fb
evroc: use base models for Whisper
2026-06-25 08:31:36 -05:00
5ae1dc5ff8
fix(azure): use base models for GPT Image
2026-06-18 11:40:46 +02:00
c02cf9ee89
refactor(zenmux): centralize model definitions and simplify provider configs
...
This refactors Zenmux model configurations by:
- Moving comprehensive model properties (e.g., limits, modalities) from `providers/zenmux/models/` to the shared `models/` directory.
- Introducing `base_model` references in `providers/zenmux/models/` files, which now primarily specify provider-specific attributes like `cost`.
- Updating parameters for `qwen3.7-plus`, `gpt-5.5-instant`, and `step-3.7-flash` during this reorganization.
2026-06-07 20:22:01 +02:00
c8a52d19c1
feat(models): add coding benchmarks and weights ( #2000 )
...
* feat(models): add coding benchmarks and weights
* feat(models): add more coding benchmarks
* feat(models): add agent benchmark scores
* feat(models): normalize benchmark metadata
2026-06-04 11:09:53 -05:00
f6c6f04367
feat(models): add model metadata ( #1974 )
...
* feat(models): add model metadata
* feat(models): rename model metadata namespaces
2026-06-03 15:13:54 -05:00
Aiden Cline