Mintplex-Labs/anything-llm
1
0
Fork 0
mirror of https://github.com/Mintplex-Labs/anything-llm.git synced 2026-07-19 22:23:50 -04:00
Code Issues 328 Packages Projects Releases 10 Wiki Activity

[GH-ISSUE #5653] [BUG]: @agent search internet exchange rate .... results in loop and timeout #5223

New Issue
Closed
opened 2026-06-05 14:52:45 -04:00 by yindo · 2 comments
yindo commented 2026-06-05 14:52:45 -04:00
Owner
Copy Link
Copy Source

Originally created by @fkam18 on GitHub (May 18, 2026).
Original GitHub issue: https://github.com/Mintplex-Labs/anything-llm/issues/5653

How are you running AnythingLLM?

AnythingLLM desktop app

What happened?

Using AppImage 1.12.1, LLM Mistral-Small-3.2-24B-Instruct-2506-Q4_K_M.gguf running under llama.cpp server on LAN; @agent search internet for today's latest exchange rate of gbp to hkd results in loop; LLM able to formulate a proper json response but anythingllm is not responding back with a local web search value etc. and ending up it says cannot search internet etc.
Using curl works and gets response back:
{"choices":[{"finish_reason":"tool_calls","index":0,"message":{"role":"assistant","content":"","tool_calls":[{"type":"function","function":{"name":"web_search","arguments":"{"query": "GBP to HKD exchange rate 2026-05-18"}"},"id":"lvtOLxtogBzy7A9C8xBqgD7u1GjUU4Gx"}]}}],"created":1779137455,"model":"Mistral-Small-3.2-24B-Instruct-2506-Q4_K_M.gguf","system_fingerprint":"b9174-59778f019","object":"chat.completion","usage":{"completion_tokens":29,"prompt_tokens":647,"total_tokens":676,"prompt_tokens_details":{"cached_tokens":0}},"id":"chatcmpl-ue8D9n7cmvJjYCrLvZyGHWJpf79fURiG","timings":{"cache_n":0,"prompt_n":647,"prompt_ms":1139.497,"prompt_per_token_ms":1.7612009273570326,"prompt_per_second":567.7943864705215,"predicted_n":29,"predicted_ms":1343.742,"predicted_per_token_ms":46.335931034482755,"predicted_per_second":21.58152383418841}}[fkam@dell ~]$

Are there known steps to reproduce?

No response

LLM Provider & Model (if applicable)

No response

Embedder Provider & Model (if applicable)

No response

Originally created by @fkam18 on GitHub (May 18, 2026). Original GitHub issue: https://github.com/Mintplex-Labs/anything-llm/issues/5653 ### How are you running AnythingLLM? AnythingLLM desktop app ### What happened? Using AppImage 1.12.1, LLM Mistral-Small-3.2-24B-Instruct-2506-Q4_K_M.gguf running under llama.cpp server on LAN; @agent search internet for today's latest exchange rate of gbp to hkd results in loop; LLM able to formulate a proper json response but anythingllm is not responding back with a local web search value etc. and ending up it says cannot search internet etc. Using curl works and gets response back: {"choices":[{"finish_reason":"tool_calls","index":0,"message":{"role":"assistant","content":"","tool_calls":[{"type":"function","function":{"name":"web_search","arguments":"{\"query\": \"GBP to HKD exchange rate 2026-05-18\"}"},"id":"lvtOLxtogBzy7A9C8xBqgD7u1GjUU4Gx"}]}}],"created":1779137455,"model":"Mistral-Small-3.2-24B-Instruct-2506-Q4_K_M.gguf","system_fingerprint":"b9174-59778f019","object":"chat.completion","usage":{"completion_tokens":29,"prompt_tokens":647,"total_tokens":676,"prompt_tokens_details":{"cached_tokens":0}},"id":"chatcmpl-ue8D9n7cmvJjYCrLvZyGHWJpf79fURiG","timings":{"cache_n":0,"prompt_n":647,"prompt_ms":1139.497,"prompt_per_token_ms":1.7612009273570326,"prompt_per_second":567.7943864705215,"predicted_n":29,"predicted_ms":1343.742,"predicted_per_token_ms":46.335931034482755,"predicted_per_second":21.58152383418841}}[fkam@dell ~]$ ### Are there known steps to reproduce? _No response_ ### LLM Provider & Model (if applicable) _No response_ ### Embedder Provider & Model (if applicable) _No response_
yindo added the possible bug label 2026-06-05 14:52:45 -04:00
yindo closed this issue 2026-06-05 14:52:45 -04:00
yindo commented 2026-06-05 14:52:46 -04:00
Author
Owner
Copy Link
Copy Source

@timothycarambat commented on GitHub (May 19, 2026):

under llama.cpp server on LAN

This would be coming from the Generic OpenAI connector, which without PROVIDER_SUPPORTS_NATIVE_TOOL_CALLING will use the fallback JSON-based tool calling approach which is almost certainly going to cause this issue.

Add this to the ENV file in the desktop storage and restart the app and the model will be much more capable now.
https://docs.anythingllm.com/installation-desktop/storage - storage/.env

PROVIDER_SUPPORTS_NATIVE_TOOL_CALLING="generic-openai"
<!-- gh-comment-id:4490922223 --> @timothycarambat commented on GitHub (May 19, 2026): > under llama.cpp server on LAN This would be coming from the Generic OpenAI connector, which without [PROVIDER_SUPPORTS_NATIVE_TOOL_CALLING](https://docs.anythingllm.com/configuration#native-tool-calling-for-llm-providers) will use the fallback JSON-based tool calling approach which is almost certainly going to cause this issue. Add this to the ENV file in the desktop storage and restart the app and the model will be much more capable now. https://docs.anythingllm.com/installation-desktop/storage - `storage/.env` ``` PROVIDER_SUPPORTS_NATIVE_TOOL_CALLING="generic-openai" ```
yindo commented 2026-06-05 14:52:46 -04:00
Author
Owner
Copy Link
Copy Source

@fkam18 commented on GitHub (May 19, 2026):

Many thks:-) Tried. Worked once.
Second to subsequent attempts it will reply with "I'm not able to browse
internet etc". Tried DuckDuckGo or my own searxng container.
Restarted the App same. Only works the first time, then subsequent attempts
negative.
Tested prompts were: @agent search CNN for latest 3 global news headlines
and display etc....
Did the same on Open WebUI and works on the searxng container but that one
I need to disable vectoring.
Tried to find any logs but can only find Event log. Not much info there.
LLM side seems normal. Mistral some 24b Q4 K V on llama.cpp 16GB GPU and
128GB system.

Cheers
Man

On Tue, May 19, 2026 at 7:42 PM Timothy Carambat @.***>
wrote:

timothycarambat left a comment (Mintplex-Labs/anything-llm#5653)
https://github.com/Mintplex-Labs/anything-llm/issues/5653#issuecomment-4490922223

under llama.cpp server on LAN

This would be coming from the Generic OpenAI connector, which without
PROVIDER_SUPPORTS_NATIVE_TOOL_CALLING
https://docs.anythingllm.com/configuration#native-tool-calling-for-llm-providers
will use the fallback JSON-based tool calling approach which is almost
certainly going to cause this issue.

Add this to the ENV file in the desktop storage and restart the app and
the model will be much more capable now.
https://docs.anythingllm.com/installation-desktop/storage - storage/.env

PROVIDER_SUPPORTS_NATIVE_TOOL_CALLING="generic-openai"

—
Reply to this email directly, view it on GitHub
https://github.com/Mintplex-Labs/anything-llm/issues/5653#issuecomment-4490922223,
or unsubscribe
https://github.com/notifications/unsubscribe-auth/AEXD3LWM7CQFX4GXB43UDOT43STI5AVCNFSM6AAAAACZDLEGS6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHM2DIOJQHEZDEMRSGM
.
You are receiving this because you authored the thread.Message ID:
@.***>

<!-- gh-comment-id:4491708141 --> @fkam18 commented on GitHub (May 19, 2026): Many thks:-) Tried. Worked once. Second to subsequent attempts it will reply with "I'm not able to browse internet etc". Tried DuckDuckGo or my own searxng container. Restarted the App same. Only works the first time, then subsequent attempts negative. Tested prompts were: @agent search CNN for latest 3 global news headlines and display etc.... Did the same on Open WebUI and works on the searxng container but that one I need to disable vectoring. Tried to find any logs but can only find Event log. Not much info there. LLM side seems normal. Mistral some 24b Q4 K V on llama.cpp 16GB GPU and 128GB system. Cheers Man On Tue, May 19, 2026 at 7:42 PM Timothy Carambat ***@***.***> wrote: > *timothycarambat* left a comment (Mintplex-Labs/anything-llm#5653) > <https://github.com/Mintplex-Labs/anything-llm/issues/5653#issuecomment-4490922223> > > under llama.cpp server on LAN > > This would be coming from the Generic OpenAI connector, which without > PROVIDER_SUPPORTS_NATIVE_TOOL_CALLING > <https://docs.anythingllm.com/configuration#native-tool-calling-for-llm-providers> > will use the fallback JSON-based tool calling approach which is almost > certainly going to cause this issue. > > Add this to the ENV file in the desktop storage and restart the app and > the model will be much more capable now. > https://docs.anythingllm.com/installation-desktop/storage - storage/.env > > PROVIDER_SUPPORTS_NATIVE_TOOL_CALLING="generic-openai" > > — > Reply to this email directly, view it on GitHub > <https://github.com/Mintplex-Labs/anything-llm/issues/5653#issuecomment-4490922223>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/AEXD3LWM7CQFX4GXB43UDOT43STI5AVCNFSM6AAAAACZDLEGS6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHM2DIOJQHEZDEMRSGM> > . > You are receiving this because you authored the thread.Message ID: > ***@***.***> >
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
Desktop
Docker
Integration Request
Integration Request
OS: Linux
OS: Mobile
OS: Windows
UI/UX
blocked
bug
bug
core-team-only
documentation
duplicate
embed-widget
enhancement
feature request
github_actions
good first issue
investigating
needs info / can't replicate
possible bug
pull-request
Mirrored from GitHub Pull Request
 question
stage: specifications
wontfix
No Label possible bug
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
yindo
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: Mintplex-Labs/anything-llm#5223
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?