Neither wins outright: Mistral: Mistral Medium 3.5 (batch) is the pick when cost matters at scale, OpenAI: GPT-5 Codex when you need to fit long inputs in one request.
| Spec | Mistral: Mistral Medium 3.5 (batch) | OpenAI: GPT-5 Codex |
|---|---|---|
| Provider | Mistral AI | OpenAI |
| Input price (per 1M tokens) | $0.75 | $1.25 |
| Output price (per 1M tokens) | $3.75 | $10.00 |
| Blended price (3:1 input/output) | $1.50 | $3.44 |
| Cheaper than … of all models we track | 31% | 19% |
| Context window | 262K tokens (~393 pages) | 400K tokens (~600 pages) |
| Larger context than … of all models | 48% | 63% |
| Max output per response | 210K tokens (~157,300 words) | 128K tokens (~96,000 words) |
| Input types | text, image, file | text, image |
| Output types | text | text |
| Free variant | No | No |
| Listed since | 2026-04-30 | 2025-09-23 |
Monthly cost at list price for five common workloads. Token counts per unit are stated so you can scale them to your own traffic. A workload that does not fit a model's context window is marked rather than priced.
| Workload (per month) | Mistral: Mistral Medium 3.5 (batch) | OpenAI: GPT-5 Codex | Difference |
|---|---|---|---|
| Customer-support chatbot 50,000 conversations — each a 6-turn conversation (1,500 in / 400 out tokens) | $131 | $294 | Mistral: Mistral Medium 3.5 (batch) saves $163 |
| RAG search over documents 20,000 questions — each a question answered from ~8 retrieved passages (6,000 in / 500 out tokens) | $127 | $250 | Mistral: Mistral Medium 3.5 (batch) saves $123 |
| Coding / tool-using agent 2,000 tasks — each a multi-step task with tool calls and file context (60,000 in / 6,000 out tokens) | $135 | $270 | Mistral: Mistral Medium 3.5 (batch) saves $135 |
| Batch summarisation 10,000 documents — each a 6,000-word report summarised to a page (8,000 in / 600 out tokens) | $82.50 | $160 | Mistral: Mistral Medium 3.5 (batch) saves $77.50 |
| Long-document analysis 500 documents — each a 150-page contract or codebase read in one pass (100,000 in / 2,000 out tokens) | $41.25 | $72.50 | Mistral: Mistral Medium 3.5 (batch) saves $31.25 |
| Capability | Mistral: Mistral Medium 3.5 (batch) | OpenAI: GPT-5 Codex | What it lets you do |
|---|---|---|---|
| Tool / function calling | ✓ Yes | ✓ Yes | call your APIs and run agent loops |
| Structured output (JSON schema) | ✓ Yes | ✓ Yes | return JSON that validates against your schema |
| Reasoning / thinking mode | ✓ Yes | ✓ Yes | spend extra tokens thinking before answering hard problems |
| Image input | ✓ Yes | ✓ Yes | read screenshots, charts and scanned pages |
| Audio input | — | — | take speech or audio directly |
| File / PDF input | ✓ Yes | — | accept documents without your own parsing step |
| Image output | — | — | generate images, not just text |
| Built-in web search | — | — | answer from live web results |
| Seed (reproducible sampling) | ✓ Yes | ✓ Yes | repeat a generation for tests and evals |
| Log-probabilities | — | — | read token confidence for classification and scoring |
From each model's published parameters and input/output types. “—” means not advertised, not proven absent.
Neither wins outright: Mistral: Mistral Medium 3.5 (batch) is the pick when cost matters at scale, OpenAI: GPT-5 Codex when you need to fit long inputs in one request. In short: Mistral: Mistral Medium 3.5 (batch) when cost matters at scale, you need file / PDF input, you generate long outputs in one go, you want the newer model; OpenAI: GPT-5 Codex when you need to fit long inputs in one request.
Mistral: Mistral Medium 3.5 (batch) is cheaper: $1.50 vs $3.44 per 1M tokens at a typical 3:1 input/output mix, 56% less. For 50,000 support conversations a month that is $131 against $294.
Mistral: Mistral Medium 3.5 (batch) costs $0.75 per 1M input tokens and $3.75 per 1M output tokens; OpenAI: GPT-5 Codex costs $1.25 input and $10.00 output. Output tokens are billed separately, so long answers shift the gap.
OpenAI: GPT-5 Codex: 400K tokens, about 600 pages, against 262K (about 393 pages) for Mistral: Mistral Medium 3.5 (batch).
Tool calling — Mistral: Mistral Medium 3.5 (batch): yes, OpenAI: GPT-5 Codex: yes. JSON-schema structured output — Mistral: Mistral Medium 3.5 (batch): yes, OpenAI: GPT-5 Codex: yes.
Yes, both accept image input.
No free variant is listed for either; both are pay-per-token.
Mistral: Mistral Medium 3.5 (batch) can write up to 210K tokens (about 157,300 words) per response; OpenAI: GPT-5 Codex up to 128K (about 96,000 words).
Mistral: Mistral Medium 3.5 (batch) was listed on 2026-04-30; OpenAI: GPT-5 Codex on 2025-09-23.
Mistral Medium 3.5 is a dense 128B instruction-following model from Mistral AI. It supports text and image inputs with text output, and is designed for agentic workflows, coding, and complex...
Full Mistral: Mistral Medium 3.5 (batch) specs →GPT-5-Codex is a specialized version of GPT-5 optimized for software engineering and coding workflows. It is designed for both interactive development sessions and long, independent execution of complex engineering tasks. The model supports building projects from scratch, feature development, debugging, large-scale refactoring, and code review. Compared to GPT-5, Codex is more steerable, adheres closely to developer instructions, and produces cleaner, higher-quality code outputs. Reasoning effort can be adjusted with the `reasoning.effort` parameter. Read the [docs here](https://openrouter.ai/docs/use-cases/reasoning-tokens#reasoning-effort-level) Codex integrates into developer environments including the CLI, IDE extensions, GitHub, and cloud tasks. It adapts reasoning effort dynamically—providing fast responses for small tasks while sustaining extended multi-hour runs for large projects. The model is trained to perform structured code reviews, catching critical flaws by reasoning over dependencies and validating behavior against tests. It also supports multimodal inputs such as images or screenshots for UI development and integrates tool use for search, dependency installation, and environment setup. Codex is intended specifically for agentic coding applications.
Full OpenAI: GPT-5 Codex specs →