Cyberax AI Playbook
cyberax.com
← All models
Model · Alibaba · Code

Qwen2.5-Coder 32B

Alibaba's 2024 open code model — 32B parameters under Apache 2.0. Superseded by Qwen3-Coder-Next, Alibaba's current open coding-agent model.

Modality
Code
License
Apache 2.0 (Open)
Parameter size
32B
Context window
131,072 tokens
Released
November 12, 2024
Last verified
October 28, 2026
Runs locally
Yes

Strengths

  • Apache 2.0 license
  • Competitive with closed coding models on HumanEval and LiveCodeBench
  • 131k context — handles full-codebase queries

Weaknesses

  • Needs 80GB GPU for full precision (or quantization)

Try it

WhereTypeNotes
Hugging Face weights Apache 2.0
Ollama local ollama run qwen2.5-coder:32b
OpenRouter hosted-api API key required

Version history

  1. Qwen2.5-Coder 32B Nov 2024 Current Deprecated
  2. Qwen 2.5 72B Sep 2024 Deprecated
  3. Qwen 2.5 7B Sep 2024 Deprecated

Official sources

Change log

  • — Deprecated — superseded by Qwen3-Coder-Next (released 2026-02-03). Weights remain on Hugging Face under Apache 2.0; no retirement date applies to self-hosted weights.
  • — Initial entry.

Related models

Kimi K2.7 Code

Moonshot AI · Code

Moonshot's coding-agent model built on Kimi K2.6 — a 1T-parameter mixture-of-experts (32B active) tuned for long, real-world software tasks, using about 30% fewer thinking tokens than K2.6. Open weights, but sized for a multi-GPU server.

Modified MIT Open weights Runs locally Mac (slow) · 512 GB+ 262k ctx
Try on: Hugging Face · Moonshot AI platform · Kimi Code
Released Jun 2026 · Verified 2026-10-28

Qwen3-Coder-Next

Alibaba · Code

Alibaba's open-weight coding-agent model for local development — an 80B mixture-of-experts with only 3B active per token, 256K context, and built to drive agent tools such as Claude Code, Cline and Qwen Code.

Apache 2.0 Open Runs locally Mac · 64 GB+ 262k ctx
Try on: Hugging Face · Ollama · OpenRouter
Released Feb 2026 · Verified 2026-10-28

Devstral Small 2 (24B)

Mistral AI · Code

Mistral's small open-weight coding-agent model — 24B parameters under Apache 2.0, 256K context, light enough for a single RTX 4090 or a 32 GB Mac. Built to explore codebases, edit many files and drive coding agents.

Apache 2.0 Open Runs locally Mac · 32 GB+ 256k ctx
Try on: Hugging Face · Ollama · Mistral Vibe
Released Dec 2025 · Verified 2026-10-28

Codestral 25.08

Mistral AI · Code

Mistral's current code-completion model, released at the end of July 2025. It is tuned for low-latency fill-in-the-middle and high-frequency code generation tasks.

Mistral Premier Proprietary 128k ctx
Try on: Mistral La Plateforme
Released Jul 2025 · Verified 2026-10-28