Add swe-bench results for qwen3-coder-next

[all-hands-bot]

Evaluation Results

Model: qwen3-coder-next
Benchmark: swe-bench
Agent Version: v1.11.1

Results

• Accuracy: 66.6%
• Total Cost: $680.26
• Average Instance Cost: $1.36
• Total Duration: 722734s (12045.6m)
• Average Instance Runtime: 1445s

---

Report Summary

• Total instances: 500
• Submitted instances: 491
• Resolved instances: 333
• Unresolved instances: 156
• Empty patch instances: 2
• Error instances: 0

Additional Metadata

• completed_instances: 489
• schema_version: 2
• unstopped_instances: 0

---

This PR was automatically created by the evaluation pipeline.

[github-actions]

📊 Progress Report

============================================================
OpenHands Index Results - Progress Report
============================================================

Target: Complete all model × benchmark pairs
  11 models × 5 benchmarks = 55 pairs
  (each pair requires all 3 metrics: score, cost_per_instance, average_runtime)

============================================================
OVERALL PROGRESS: ⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛ 100.0%
  Complete: 55 / 55 pairs
============================================================

❌ Schema Validation

============================================================
Schema Validation Report
============================================================

Results directory: /home/runner/work/openhands-index-results/openhands-index-results/results
Files validated: 30
  Passed: 29
  Failed: 1

Errors:
  - /home/runner/work/openhands-index-results/openhands-index-results/results/qwen3-coder-next/metadata.json:   • Field 'model': Input should be 'claude-4.6-opus', 'claude-4.5-opus', 'claude-4.5-sonnet', 'gemini-3-pro', 'gemini-3-flash', 'glm-4.7', 'gpt-5.2', 'gpt-5.2-codex', 'kimi-k2-thinking', 'kimi-k2.5', 'minimax-m2.1', 'deepseek-v3.2-reasoner', 'qwen-3-coder' or 'nemotron-3-nano' (got: 'qwen3-coder-next')

============================================================
VALIDATION FAILED
============================================================

This report measures progress towards the 3D array goal (benchmarks × models × metrics) as described in #2.

[juanmichelini]

@OpenHands search the web and add fields for "country", "release_date", "parameter_count_b", "active_parameter_count_b" show your work add a comment in the PR with links proving each of the fields.

[openhands-ai]

I'm on it! juanmichelini can track my progress at all-hands.dev

[juanmichelini]

Sources for Added Metadata Fields

I have added the following fields to qwen3-coder-next/metadata.json:

Field	Value	Source
country	cn	Alibaba is a Chinese company headquartered in Hangzhou, China
release_date	2026-02-02	Qwen Official Blog - Shows "2026/02/02" as the publication date
parameter_count_b	80	Hugging Face Model Card - "Model size: 80B params"
active_parameter_count_b	3	Hugging Face Model Card - "80B total parameters and only 3B activated per token"

Additional Supporting Sources:

• NewsBreak Article - "Released February 3 under an Apache 2.0 license", "80 billion parameters that activates just 3 billion per forward pass"
• VentureBeat - "80 billion total parameters, it utilizes an ultra-sparse Mixture-of-Experts (MoE) architecture that activates only 3 billion parameters per forward pass"
• Dev.to Guide - "Total Parameters: 80B, Activated Parameters: 3B (per inference)"

[openhands-ai]

Looks like there are a few issues preventing this PR from being merged!

• GitHub Actions are failing:
  • Measure Progress

If you'd like me to help, just leave a comment, like

@OpenHands please fix the failing actions on PR #539 at branch `eval/qwen3-coder-next/swe-bench-20260209-234004`

Feel free to include any additional details that might help me get this PR into a better state.

_{^{You can manage your notification settings}}

[openhands-ai]

Summary

I have completed the requested task to add metadata fields for the qwen3-coder-next model. Here's what was done:

Changes Made

Updated results/qwen3-coder-next/metadata.json with the following fields:

Field	Value	Source
country	cn	Alibaba is a Chinese company headquartered in Hangzhou, China
release_date	2026-02-02	Qwen Official Blog
parameter_count_b	80	Hugging Face Model Card
active_parameter_count_b	3	Hugging Face Model Card

Checklist

• Added country field
• Added release_date field
• Added parameter_count_b field
• Added active_parameter_count_b field
• Committed and pushed changes to the PR branch
• Added a comment on PR Add swe-bench results for qwen3-coder-next #539 with links proving each field's source

PR Comment

A detailed comment with source links has been added to PR #539: #539 (comment)

The comment includes:

• Primary sources (Qwen Official Blog, Hugging Face Model Card)
• Additional supporting sources (NewsBreak, VentureBeat, Dev.to)

View full conversation

[juanmichelini]

LGTM

[juanmichelini]

@OpenHands fix the validation errors!

[openhands-ai]

I'm on it! juanmichelini can track my progress at all-hands.dev

[juanmichelini]

https://github.com/OpenHands fix the validation errors!