Model crosswalk
Side-by-side on price, capability and workload — three-way comparison.
Codestral 22b
vs
Qwen 2.5 Coder 32b Instruct
vs
Qwen 3 32b Instruct
Codestral 22bA
Codestral 22b
Cheapest provider—
$/1M input—
$/1M output—
Qwen 2.5 Coder 32b InstructB
Qwen 2.5 Coder 32b Instruct
Cheapest provider—
$/1M input—
$/1M output—
Qwen 3 32b InstructC
Qwen 3 32b Instruct
Cheapest provider—
$/1M input—
$/1M output—
Specs and cheapest providers
| Spec | Codestral 22b | Qwen 2.5 Coder 32b Instruct | Qwen 3 32b Instruct |
|---|---|---|---|
| Parameters | — | — | — |
| Context window | — | — | — |
| License | — | — | — |
| Released | — | — | — |
| Cheapest provider | |||
| Provider | — | — | — |
| Input / 1M tokens | — | — | — |
| Output / 1M tokens | — | — | — |
Benchmark comparison
No benchmark data available yet.
Editor's take
This comparison pits a licensed coding specialist against two commercially permissive 32B-class models — one code-focused and one general-purpose with multilingual breadth.
Codestral 22B is Mistral AI's first code-specialist model, a 22 billion parameter dense transformer released May 2024. It targets 80-plus programming languages with a 32K context window and performs competitively on HumanEval against same-era coding peers. The Mistral Research License is the standing concern: production commercial deployment requires a direct agreement with Mistral, which adds a legal overhead that many teams discover late in evaluation cycles.
Qwen 2.5 Coder 32B Instruct, released November 2024, is Alibaba's purpose-built coding model at 32 billion parameters with a 131K context window and support for 92 programming languages. On LiveCodeBench and MultiPL-E it benchmarks alongside DeepSeek Coder V2 — one of the stronger showings in the sub-frontier coding tier. The Qwen license permits commercial deployment, and hosted pricing is available across multiple inference providers. For API-scale code generation, this is the more straightforward production path of the three.
Qwen 3 32B Instruct, from Alibaba's Qwen 3 series, is a general-purpose instruction model at 32 billion parameters with a 131K context window and strong multilingual performance across CJK languages and Arabic. It delivers roughly 85 percent of Qwen 3 72B benchmark results at a proportionally lower price point. Coding performance is competent but not code-specialist tuned — it handles mixed-language tasks and multi-domain pipelines that Qwen 2.5 Coder 32B's narrow specialization misses. The Qwen license covers commercial use.
Pick Codestral 22B for internal, non-commercial tooling where the research license fits. Pick Qwen 2.5 Coder 32B when your workload is primarily code generation and you want the sharpest coding benchmark scores in this size tier. Pick Qwen 3 32B Instruct when your pipeline mixes code with multilingual natural language tasks, or when you need broader capability coverage from a single model.
Compare two at a time
Frequently asked questions
- How does Codestral 22b compare to Qwen 2.5 Coder 32b Instruct and Qwen 3 32b Instruct on price?
- Use the table above to compare input and output prices per 1M tokens across the cheapest available providers for each model.
- Which model is best for coding: Codestral 22b, Qwen 2.5 Coder 32b Instruct, or Qwen 3 32b Instruct?
- HumanEval and other code benchmarks are shown in the table. For production code tasks, also consider context window size and provider latency.
- What is the context window for Codestral 22b, Qwen 2.5 Coder 32b Instruct, and Qwen 3 32b Instruct?
- Context window sizes are listed in the Specs row of the comparison table above.
Full model details