b59eb6cbfb
feat: Add Anthropic web search and web fetch tools support
...
Implements comprehensive support for Anthropic's web search (web_search_20250305) and web fetch (web_fetch_20250910) tools.
- Add llmVndAntWebSearch and llmVndAntWebFetch parameters with ['auto', 'off'] options
- Enable tools for Claude 4.5, 4.1, 4, 3.7, 3.5 Sonnet/Haiku/Opus models (including thinking variants)
- Inject web_search_20250305 and web_fetch_20250910 tools based on parameter values
- Configure web search with max_uses=5 for progressive searches
- Configure web fetch with max_uses=5 and citations enabled
- Add dynamic beta header injection for web fetch (web-fetch-2025-09-10)
- Add UI controls in model settings for easy parameter configuration
Parser already supports web_search_tool_result and web_fetch_tool_result blocks (no changes needed).
🤖 Generated with [Claude Code](https://claude.com/claude-code )
Co-Authored-By: Enrico Ros <enricoros@users.noreply.github.com >
2025-10-18 07:34:51 +00:00
a75a31ff04
AIX: Anthropic: Parser: improve
2025-10-18 00:05:09 -07:00
a0f97e9cd8
AIX: Anthropic: major protocol update
2025-10-17 23:40:31 -07:00
fe6e7245de
AIX: Anthropic: remove vnd.ant.tools for upgrade soon
2025-10-17 23:13:29 -07:00
0de37e337b
LLMs: OpenAI: added GPT-5'class models the no-stream option for unverified orgs
2025-10-16 11:37:27 -07:00
9bbc2a2e00
fix: Add llmForceNoStream parameter to OpenAI models with Org ID streaming issues
...
- Added parameter to GPT-5, GPT-5 Pro, GPT-5 Codex, and GPT-5 Mini
- These models require Organization ID verification for streaming
- Kept existing parameter on o3 and o3-pro models
- Did not modify GPT-5 Nano, GPT-5 Chat Latest, GPT-4.1, or GPT-4o models which work fine
Fixes #847
Co-authored-by: Enrico Ros <enricoros@users.noreply.github.com >
2025-10-16 18:09:53 +00:00
3e4e6b2f4b
BackupRestore: fix a potential removal issue
2025-10-16 10:28:37 -07:00
e6a65bdf8e
BackupRestore: improve messaging
2025-10-16 10:05:05 -07:00
740d76c15c
LLMs: OpenAI: add gpt-5-search-api
2025-10-15 19:36:38 -07:00
ca4d21d4b8
LLMs: Gemini: improve logs
2025-10-15 19:19:30 -07:00
674c9c8c25
LLMs: Ollama: update and remove all descriptions (not used)
2025-10-15 17:29:02 -07:00
98a3e7e185
LLMs: Alibaba: update: remove any model information and jut return the sorted list from the API
2025-10-15 17:29:02 -07:00
ee00c53ada
LLMs: Perplexity: update
2025-10-15 17:29:02 -07:00
0553f64fe8
LLMs: xAI: update
2025-10-15 17:29:02 -07:00
ff06f6f04c
LLMs: Groq: update
2025-10-15 17:29:02 -07:00
3f45617e06
LLMs: Deepseek: update
2025-10-15 17:29:02 -07:00
9d93c8c55a
LLMs: Mistral: update all
2025-10-15 17:29:02 -07:00
73eaf740db
LLMs: Anthropic: Haku 4.5 (Thinking and not)
2025-10-15 17:29:02 -07:00
48426d5022
AIX: OpenAI: [Mistral, 2025-10-15]: non-standard delta.content for thinking blocks
2025-10-15 17:29:02 -07:00
73f8488d22
AIX: enable Abort on GET
2025-10-15 12:58:27 -07:00
92e371837d
Modals: fullscreen options on desktop
2025-10-14 17:54:40 -07:00
eb39db9974
KB: structure
2025-10-14 17:31:40 -07:00
087e6e2eaf
AIX: Client: decimator cleanups
2025-10-14 02:52:28 -07:00
295d91b310
AIX: Client: improve disconnect message.
2025-10-14 02:50:50 -07:00
f75bcb78d7
FIX: AIX: Anthropic: hotfix: disable thinking with forced tool use (policy of: any, tool)
2025-10-14 02:48:22 -07:00
dc75136131
LLMs: OpenAI: flush
2025-10-14 01:15:28 -07:00
57c43b3c4e
LLMs: OpenAI: reduce model desc size
2025-10-14 01:12:03 -07:00
4c5b7677e6
LLMs: nit
2025-10-14 01:12:03 -07:00
43890150e5
AIX: OpenAI: fix NS function parsing
2025-10-14 01:11:19 -07:00
ef1f412019
LLMs: OpenAI: slimmer
2025-10-13 18:51:14 -07:00
1249efb53b
LLMs: xAI: add the latest groks. Fixes #837
2025-10-13 18:50:21 -07:00
8bc81e45ce
Models: UI hide hidden by default
2025-10-13 17:52:23 -07:00
3269e10da9
LLMs: OpenAI: enable Audio output models
2025-10-13 17:10:23 -07:00
53a57fd7ff
AIX: OpenAI: support Audio output models S/NS
2025-10-13 17:10:09 -07:00
dbbf25c3af
LLMs: new model description builder
2025-10-13 17:09:17 -07:00
a2ff00f53b
LLMs: mandatory fallbacks
2025-10-13 13:49:08 -07:00
8221444308
LLMs: Gemini: enable search
2025-10-13 13:24:31 -07:00
7cd94b3163
*AIX: Gemini: Google Search support
2025-10-13 13:24:16 -07:00
52cdf7da4e
AIX: Hosted tools - bits of cleanup
2025-10-13 12:46:02 -07:00
6ff010ae0e
AIX: Gemini: remove the old google_search_retrieval
2025-10-13 12:05:19 -07:00
6d81150975
AIX: Gemini: parse also finish message and url grounding
2025-10-13 11:55:44 -07:00
0fdcc4c64d
LLMs: Gemini: remove 4:5 and 5:4 support, not in the API
2025-10-13 11:52:30 -07:00
f272c9cb12
AIX: Gemini: expand to all finish reasons
2025-10-13 11:36:11 -07:00
5354f83736
AIX: OpenAI: Responses: uniform S/NS parsing
2025-10-13 11:26:57 -07:00
f4b2f36ac0
AIX: OpenAI: Responses: handle streaming stop reasons
2025-10-13 10:49:59 -07:00
5fca834c20
AIX: OpenAI: Responses: handle max_tokens
2025-10-13 10:49:59 -07:00
fff48335ae
AIX: OpenAI: skip also markdown restoration when custom tools with restrictive policies are applied
2025-10-13 10:49:59 -07:00
f39a1825cf
AIX: OpenAI: skip hosted tools when custom tools with restrictive policies are applied
2025-10-13 10:17:21 -07:00
c1b10405a5
AIX: Anthropic: document context_window_exceeded
2025-10-13 10:17:21 -07:00
37ba583cf2
AIX: Anthropic: parse new stop reasons
2025-10-13 10:17:21 -07:00
claude[bot]