Opencode + Qwen gives JSON output instead of writing to disk

Description

I'm using OpenCode 1.15.11 with Qwen2.5-code:7b running with ollama on my desktop.

I asked for a simple python script, saved into a file. Instead of creating the file on disk, OpenCode returned the JSON for the write command.

I'm using the build agent and here is my ./.config/opencode/opencode.jsonc. I understand that I have given all the necessary permissions to the model

{
    "$schema": "https://opencode.ai/config.json",
    "permission": {
	"edit": "ask",
	"bash": "ask",
	"webfetch": "allow",
	"read": "ask",
	"write": "ask"
    },
   "provider": {
       "ollama": {
	   "name": "ollama",
	   "npm": "@ai-sdk/openai-compatible",
           "options": {
               "baseURL": "http://localhost:11434/v1"
           },
           "models": {
      	       "qwen2.5:7b": { "name": "Qwen2.5 7b" },
      	       "qwen2.5-coder:7b": { "name": "Qwen2.5 Code 7b" },
	       "qwen3.5:9b": { "name": "Qwen3.5 9b" }
           }
       }
   }
}

Plugins

No response

OpenCode version

1.15.11

Steps to reproduce

Install ollama and Qwen2.5-code:7b (using 16k context size)
Instruct opencode/Qwen2.5 to create the file

Screenshot and/or share link

No response

Operating System

Ubuntu

Terminal

No response

This issue might be a duplicate of existing issues. Please check:

Ollama (qwen2.5-coder): tool calls (edit/write) show as executed but no files are created/modified #7030: Ollama (qwen2.5-coder): tool calls (edit/write) show as executed but no files are created/modified — same model, same symptom (JSON tool payload printed but files not written to disk)
local model keeps returning json and doesn't create any files #18454: local model keeps returning json and doesn't create any files — same root symptom with Ollama + Qwen
not compatible to ollama/qwen2.5-coder:7b #10402: not compatible to ollama/qwen2.5-coder:7b — same model/provider combination causing tool call issues

I have this issue too. It happens with windows

u can sove this issue by Increase context length

I'm using a bigger context length and it doesn't work on each of the following models, context size: 32k. Same problem of the OP. Version 1.15.10

{
  "$schema": "https://opencode.ai/config.json",
  "model": "ollama/qwen2.5-coder:7b-32k",
  "provider": {
    "ollama": {
      "npm": "@ai-sdk/openai-compatible",
      "name": "Ollama (Local Tool-Capable)",
      "options": {
        "baseURL": "http://127.0.0.1:11434/v1"
      },
      "models": {
        "qwen2.5-coder:7b-32k": {
          "name": "Qwen 2.5 Coder 7B (32k Context)"
        },
        "qwen2.5-coder:7b": {
          "name": "Qwen 2.5 Coder 7B (Default Context)"
        },
        "llama3.1:8b-32k": {
          "name": "Llama 3.1 8B (32k Context)"
        },
        "llama3.1:8b": {
          "name": "Llama 3.1 8B (Default Context)"
        }
      }
    }
  }
}

Update:

adding

"permission": {
	"edit": "ask",
	"bash": "ask",
	"webfetch": "allow",
	"read": "ask",
	"write": "ask"
    },

that was missing, llama3.1:8b-32k and lama3.1:8b seem to work. You need to specify the full path location where to write files but at least seems to produce something. Qwen still completely broken

With working llama 3.1 what i'm noticing is that it is not able to write in correct locations:

sh-5.3$ pwd
/tmp/we
sh-5.3$ cat ../test.txt 
test

I tested with both default and 16k context length and none worked

I tested with both default and 16k context length and none worked

I found something working:

google/gemma-4-E4B

But with continue.dev visual studio code extension, not using opencode. Seems to have a lot of problems with small models

Upgraded to opencode v1.15.13

Still, local Qwen model will not write to disk.

Changing model to DeepSeek V4 Flash Free (from OpenCode Zen), it works as expected and the file is writen.

Looks like it's a specific problem from Qwen?

@Virgula0

Using @Virgula0 suggestion (Gemma 4 E4B) it works for me as well, but I could not get it to work with Qwen 2.5B Instruct.

Maybe a heads-up for someone else: I had to create a new session in OpenCode for it to work

I also have the same issue and couldn't find any solution.

When using qwen2.5-coder:32b, the output is:

opencode run "/no_think generate a todo.md file with the contents 'hello world, from qwen3' in it"

build · qwen2.5-coder:32b

write: {"content": "hello world, from qwen3", "filePath": "/tmp/test/todo.md"}

And no file is written.

With Gemma4, the same works nicely:

opencode run "/no_think generate a todo.md file with the contents 'hello world, from gemma' in it"

build · gemma4:31b

✱ Glob "todo.md" 0 matches

← Write todo.md
Wrote file successfully.

Done.

My opencode.json looks like this (with the line "model": "ollama/qwen2.5-coder:32b", being changed when gemma4 is used):

{
  "$schema": "https://opencode.ai/config.json",
  "model": "ollama/qwen2.5-coder:32b",
  "provider": {
    "ollama": {
      "models": {
        "qwen2.5-coder:32b": {
          "name": "qwen2.5-coder"
        },
        "gemma4:31b": {
          "_launch": true,
          "name": "gemma4"
        }
      },
      "name": "ollama",
      "npm": "@ai-sdk/openai",
      "options": {
        "baseURL": "http://172.17.0.1:11434/v1",
        "apiKey": "anything"
      }
    }
  }
}

I used 1.18.4 and ollama 0.32.1, so everything should be up to date. If there is any workaround, I'd be happy to know about this.

EDIT qwen3.6:27b works, since this is much newer and has comparable size, it probably makes sense to just switch to this model.

To stay organized issues are automatically closed after 60 days of no activity. If the issue is still relevant please open a new one.

I'm also having the same problem with any of the qwen models I've tried.

Basically none of the invocation commands from qwen work in OpenCode.

qwen2.5-coder:7b responses are different from qwen3.5:9b responses, but neither work.

This problem is also not isolated to OpenCode. I tried running Claude-Code locally and connecting it to Ollama -> qwen2.5-coder:7b and basically it has the same problem as OpenCode.

As qwen2.5 and qwen3.5 responses seem to be different, the only way this could be resolved would be for opencode to be able to be configured to handle the difference response formats for these 2 LLMs, and as they are not the latest version of qwen, I doubt they would want to devote effort to support them :-(

As Gemma4 seems to have worked for some people, I'll give that a go

Navigation Menu

Repository navigation

Opencode + Qwen gives JSON output instead of writing to disk #29757

Description

Plugins

OpenCode version

Steps to reproduce

Screenshot and/or share link

Operating System

Terminal

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Opencode + Qwen gives JSON output instead of writing to disk #29757

Description

Description

Plugins

OpenCode version

Steps to reproduce

Screenshot and/or share link

Operating System

Terminal

Activity

github-actions commented on May 28, 2026

QQi44 commented on May 30, 2026

QQi44 commented on May 30, 2026

Virgula0 commented on May 31, 2026

dvictori commented on Jun 1, 2026

Virgula0 commented on Jun 1, 2026

dvictori commented on Jun 3, 2026

brtmax commented on Jun 16, 2026

DaGeRe commented on Jul 23, 2026

github-actions commented on Sep 23, 2026

rogerclarkmelbourne commented on Oct 31, 2026

Metadata

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions