Last checked 30 Oct 2026
DeepSeek V4 Pro vs Mistral Large 4
DeepSeek V4 Pro is billed by DeepSeek. Mistral Large 4 is billed by Mistral. A pair is two models from different providers whose higher published input rate is at most 1.5 times the lower one, and whose higher published output rate is at most 1.5 times the lower one. Closer rates are listed first. Each model is in at most 2 pairs, and the list stops at 12. The figures below are the rates stored on those two rows, the differences between them, and the cost of one sample workload. This page does not score quality.
Stored rates
DeepSeek V4 Pro (DeepSeek, Off-peak) and Mistral Large 4 (Mistral, Standard). Provider pages: DeepSeek and Mistral.
| Field | DeepSeek V4 Pro | Mistral Large 4 |
|---|---|---|
| Input / 1M | $0.66 | $0.70 |
| Cached input / 1M | $0.022 | $0.07 |
| Output / 1M | $1.98 | $2.10 |
| Rate | Off-peak | Standard |
| Context | 1M | 1M |
| Max output | 384k | n/a |
| Status | Active | Preview |
| Modalities | text → text | n/a |
| Released | n/a | 6 Oct 2026 |
| Checked | 30 Oct 2026 | 30 Oct 2026 |
| Per request | $0.002904 | $0.00308 |
| Per month | $87.12 | $92.40 |
Sample workload
The sample is the calculator's starting workload: 2,000 input tokens and 800 output tokens per request, 1,000 requests a day. A month in the calculator is 30 days, so that is 30,000 requests. The cost uses the stored input and output rates only. Cache hits, batch discounts, long-context surcharges, peak hours, and prices scheduled after an introductory window are not included. The rate label and the notes below name those other published rates.
DeepSeek V4 Pro is $0.002904 per request and $87.12 per month ($39.60 input and $47.52 output over the month). Mistral Large 4 is $0.00308 per request and $92.40 per month ($42.00 input and $50.40 output over the month).
DeepSeek V4 Pro costs $5.28 less per month than Mistral Large 4 on this sample, using the stored rates above.
Open DeepSeek V4 Pro in the calculatorOpen Mistral Large 4 in the calculator
Notes from the sources
DeepSeek V4 Pro
Off-peak
Off-peak cache-miss input, off-peak cache-hit, and off-peak output. Peak is double: $1.32 input, $0.044 cache hit, and $3.96 output per 1M tokens, during the same peak hours as DeepSeek Flash. Context is published as 1M and max output as 384K. Vision is marked not supported. Model version on the page is DeepSeek-V4-Pro-0813. A separate release day was not on this page.
Mistral Large 4
Standard
Standard USD rates from the API pricing table: $0.70 input, $0.07 cached input, and $2.10 output per 1M tokens. On 29 October 2026 Mistral released open weights for an improved Mistral Large 4 (Le Chonk) under the Mistral Ecosystem Open-Weight License, a modified MIT license that requires a separate commercial license for Model-as-a-Service, and priced the API at those rates; the previous ledger rate was $1.36 input, $0.14 cached input, and $4.18 output. The model card lists a 1M context window and Public Preview (v26.10). Max output is not published.