163ef9296e
HBWA: note
2025-11-13 01:48:18 -08:00
fa042f7d68
AIX: prenorm
2025-11-12 22:53:50 -08:00
8a11040dde
optimization: faster 'hasKeys' Object non-emptiness check (avoids allocation)
2025-11-12 22:40:46 -08:00
a88971d557
Models: Vendor (service) selection: vastly improve
2025-11-12 17:45:34 -08:00
5867e5fcc5
OpenRouter: more config beauty
2025-11-12 17:32:43 -08:00
20e587d6d3
OpenRouter: more config cleanup
2025-11-12 16:58:14 -08:00
6bfa8471cd
Models Modal: option for fullscreen
2025-11-12 16:01:14 -08:00
5c10bce2f4
ModelsList: disable the large popups
2025-11-12 15:53:52 -08:00
90c27e0e74
LLMs: add displayGroup
2025-11-12 15:31:32 -08:00
b5eac0d907
OpenRouter: more config improvement
2025-11-12 15:20:30 -08:00
a1c0d30a06
LLMs: GPT-5 optimistic updates (coming later this week)
2025-11-12 15:10:50 -08:00
63c9f65040
Merge remote-tracking branch 'opensource/claude/issue-879-20251112-2245'
2025-11-12 15:05:01 -08:00
f58a066bff
OpenRouter: improve config
2025-11-12 15:02:21 -08:00
6695973035
tRPC: raise the correct server-side exceptions
2025-11-12 14:57:44 -08:00
3dc28635f4
Add support for GPT-5.1 models
...
- Add gpt-5.1-2025-11-12 (GPT-5.1 Thinking) with adaptive reasoning
- Add gpt-5.1-chat-latest (GPT-5.1 Instant) with adaptive reasoning
- Both models include full feature set: chat, vision, function calling, JSON, prompt caching, reasoning, web search
- Pricing set to match GPT-5 (to be updated when official pricing is announced)
- Added models to manual ordering list for proper UI sorting
Co-authored-by: Enrico Ros <enricoros@users.noreply.github.com >
2025-11-12 22:47:51 +00:00
0bde01a85f
Llms: vector ops
2025-11-12 14:43:26 -08:00
8228a76875
LLMs: LocalAI enumeration less verbose
2025-11-11 18:40:52 -08:00
46b370a2e3
AIX: OpenAI ChatGenerate: fixup for malformed NS objects - best-effort
2025-11-11 18:37:34 -08:00
bd71d64db3
LLMs: OpenRouter: fix context window max_tokens issue
2025-11-11 18:09:50 -08:00
9d4baf827c
LLMs: OpenRouter: auto-detection of capabilities (i/o modalities, features, etc). Thanks OpenRouter, you're the best!
2025-11-11 17:47:31 -08:00
d6843d7fcf
AIX: OpenRouter: write/parse the new reasoning request / reasoning_details response
...
Removes older system(s) from OpenRouter.
2025-11-11 17:46:58 -08:00
babb1dd962
LLMs: OpenRouter: parser
2025-11-11 16:46:13 -08:00
09a38c0e4b
UserInputChecklistComponent: fix crash
2025-11-11 12:30:02 -08:00
645b8fb9cd
LLMs: OpenRouter: fix 400
2025-11-11 12:14:37 -08:00
541588948c
LLMs: OpenRouter: add future PKCE
2025-11-11 12:13:09 -08:00
9e50286c66
LLMs: Moonshot: remove images from the request for those models
2025-11-10 02:00:42 -08:00
85146d8af0
LLMs: Moonshot: native search parameter (disabled here and in the model - upstream not good enough)
2025-11-09 20:16:39 -08:00
9612572f07
AIX: Moonshot: native search dispatch/parse support
2025-11-09 20:15:02 -08:00
4bb1dddf4d
ApproximateCosts: only display on Extra
2025-11-09 20:15:02 -08:00
b066a86962
AIX: fix exhaustive check
2025-11-09 20:15:02 -08:00
6086455782
LLMs/AIX: Moonshot AI support
2025-11-09 20:15:02 -08:00
9020b3cbad
AIX: Fix Mistral's response breakage in NS mode - Fixes #873 .
2025-11-09 17:23:07 -08:00
5822dea270
AIX: Further breakdown logging levels - with Retry entry points.
2025-11-09 17:14:55 -08:00
c445f59664
AIX: Fix Logging, with warn only for real server-side solvable issues. Fixes #872 . Fixes #874.Fixes #875
2025-11-09 17:07:14 -08:00
737e4cb4f9
AIX - fix Gemini parser. Fixes #871
2025-11-09 16:31:52 -08:00
dba7368d01
Rename platform format
2025-11-07 23:35:29 -08:00
3e46f99e14
AIX: Error correction: client-side reporting
2025-11-07 22:45:11 -08:00
e0cc552b8d
AIX: Error correction: operation reporting (Anthropic only for now, but generic)
2025-11-07 22:45:11 -08:00
6b5be403af
AIX: Error correction: dispatch reporting
2025-11-07 22:45:11 -08:00
b7de19b020
AIX: turn on operation-retrier for Anthropic's error-in-stream. Fixes #869
2025-11-07 12:03:16 -08:00
77cd659b39
AIX: support operation-level retrier with reassembly wipe #869
2025-11-07 12:03:12 -08:00
f464a9efdf
LLMs: listModels: openAI, deprio 401. Fixes #870
2025-11-07 11:42:22 -08:00
7ec4290582
AIX: Anthropic: retriable errors: extended parsing
2025-11-07 11:04:21 -08:00
3f887a1d3a
AIX: Gemini: fix other proxy errors. Fixes #868
2025-11-07 10:50:25 -08:00
ffd76dc587
*Image Captioning with a dedicated (configurable) model. Fixes #862
2025-11-05 14:15:50 -08:00
17f03806d0
AIX: rename DMessage gen to throw
2025-11-05 07:07:05 -08:00
b6aba0efa4
AIX: type change to DMessageGuts for correctness
2025-11-04 17:01:26 -08:00
f5470aca5d
AIX: OpenAI Responses: remove patch for deep-research
2025-11-02 20:12:07 -08:00
c26af97fe7
LLMs: OpenAI: deep research with standard params path
2025-11-02 20:11:17 -08:00
da19ef42f5
AIX: OpenAI Responses: use the web_search (non preview) tool
2025-11-01 17:19:57 -07:00
Enrico Ros