Models / Mistral
Mistral Large 3 2512
by Mistral ·
mistralai/mistral-large-2512Mistral Large 3 2512 is Mistral’s most capable model to date, featuring a sparse mixture-of-experts architecture with 41B active parameters (675B total), and released under the Apache 2.0 license.
tool-usevisionreleased 2025-12-01
Legal score · raw
-
not yet benchmarked
Context
262K
max output 210K
Input
$0.63
per 1M tokens
Output
$1.88
per 1M tokens
Suite cost
-
no run recorded
Prompt caching
Grounded answers re-send the same retrieved authority every turn, so a cached read is what a long matter actually costs. Prices are per 1M tokens, from the provider’s own catalog.
Cache read
$0.06
10x cheaper than uncached
Uncached input
$0.63
per 1M tokens
Capabilities
Modalities
in textimagefile
out text
Request parameters it accepts
frequency_penaltymax_tokenspresence_penaltyresponse_formatseedstopstructured_outputstemperaturetool_choicetoolstop_p
Benchmark results
This model has not been run through the suite yet. See Humanity's Last Lawsuit for the benchmark models are ranked on.
Measured by DocketBuster
These numbers come from DocketBuster's own legal battery, not from DocketRouter's suite. Latest run per metric, with n and a 95% Wilson interval where the source reports one. See docketbuster.com/benchmarks.
| Metric | Value | n | Interval | Measured |
|---|---|---|---|---|
| Statute pinpoint, exact section (no retrieval) | 20.7% (62/300) | 300 | 95% CI 16.5% to 25.6% | 2026-08-22 |
| Statute pinpoint, exact section (with DocketBuster retrieval) | 77.9% (239/307) | 307 | 95% CI 72.9% to 82.1% | 2026-08-24 |
| Say-nothing rate (declines to bluff when the answer is not in the record) | 100.0% | 189 | 95% CI 98.0% to 100.0% | 2026-08-22 |
| Abstained on statute pinpoint | 5.9% (18/307) | 307 | count, no interval reported | 2026-08-24 |
| Coaching quality (rubric, 0 to 8) | 6.92 / 8 | - | rubric mean, no interval reported | 2026-08-23 |