GPQA DiamondVerified
85.7%
GPQA Diamond (AA run)
Models / Google
Gemma 4 31B IT
Gemma · Released 2026-03-01
Google's March 2026 open-weights Gemma 4 instruct model. SAGE 55% on Vals (#2). Small enough to self-host on a single GPU cluster.
Self-hosted Google open weights for privacy-sensitive workflows that can't call cloud APIs.
Composite53
Context131K
Input / 1M—
Output / 1M—
Knowledge cutoffopen
Output speed35 t/s
Statussolid
Family profile
Best published score in each covered benchmark family.
61/100 avg
- Reasoning
- 86
- Coding
- 46
- Agentic
- 43
- Long context
- 70
4 tested benchmark families
Benchmark placements
Per-benchmark positions
Where Gemma 4 31B IT places on each public benchmark source that publishes a row. Rank counts every model with a latest row in the same test — not a universal quality score.
| Benchmark | Family | Rank | Score | Source | Date |
|---|---|---|---|---|---|
| SciCode | Coding | #64/ 97 | 46 | Artificial Analysis | 2026-09-28 |
| Terminal-Bench | Agentic | #87/ 167 | 43 | Artificial Analysis | 2026-09-28 |
| GPQA Diamond | Reasoning | #90/ 184 | 86 | Artificial Analysis | 2026-09-28 |
| Humanity's Last Exam | Reasoning | #115/ 186 | 24 | Artificial Analysis | 2026-09-28 |
| AA-LCR | Long context | #125/ 184 | 70 | Artificial Analysis | 2026-09-28 |
| AA output speed | Performance | #141/ 195 | Artificial Analysis | 2026-10-12 |
Family context
Strongest rows in each family
The three highest-scoring models with pages in each capability family. Where Gemma 4 31B IT shows up, it's highlighted.
ReasoningGemma 4 31B IT · 86
- 196
- 295
- 395
Math
- 152
- 252
- 350
CodingGemma 4 31B IT · 46
- 174
- 273
- 370
AgenticGemma 4 31B IT · 43
- 191
- 289
- 389
Long contextGemma 4 31B IT · 70
- 189
- 286
- 385
Newest receipts
6 recentLast 120 days
Benchmark rows added to the public ledger for Gemma 4 31B IT in the last 120 days. Older rows live in the full table below.
| Benchmark | Family | Score | Source | Days ago |
|---|---|---|---|---|
| AA output speed | Performance | 35 t/s | Artificial Analysis | 10 |
| AA-LCR | Long context | 69.67% | Artificial Analysis | 24 |
| GPQA Diamond | Reasoning | 85.7% | Artificial Analysis | 24 |
| Humanity's Last Exam | Reasoning | 23.6% | Artificial Analysis | 24 |
| SciCode | Coding | 45.5% | Artificial Analysis | 24 |
| Terminal-Bench | Agentic | 43.45% | Artificial Analysis | 24 |
Family coverage
Best score per family
A high composite that hides a weak family is a trap. These bars surface the families where this model hasn't been publicly tested, and where it leads.
Price vs. performance
Where Gemma 4 31B IT sits on the scatter
Hover any dot for name, score, and input price. Keyboard: tab through the top twelve, or use the ranking below.
AnthropicOpenAIGoogleAlibabaMetaZhipuxAIDeepSeekMiniMaxXiaomiMoonshotOtherMeituanMistralNVIDIACohere
Full benchmark ledger
Every catalog benchmark for Gemma 4 31B IT. Scores link to the original source; gaps mean no public row exists yet.
6 of 13 catalog benchmarks have a sourced row for Gemma 4 31B IT.
Reasoning
Humanity's Last ExamVerified
23.6%
HLE (AA run)
Coding
SciCodeVerified
45.5%
SciCode (AA run)
Agentic
Terminal-BenchVerified
43.45%
Terminal-Bench v2.1 (AA run)
Long context
AA-LCRVerified
69.67%
AA-LCR (AA run)
Performance
AA output speedVerified
35 t/s
Median output tokens/s (1k prompt, default provider)
Sources
Where the numbers come from
- Gemma docs· checked 2026-06-09
Questions this model page answers
Every answer below is assembled from the dated fields on this page. Nothing is written separately for search.
What is Gemma 4 31B IT?
Gemma 4 31B IT is a Google model: Google's March 2026 open-weights Gemma 4 instruct model. SAGE 55% on Vals (#2). Small enough to self-host on a single GPU cluster.
What is Gemma 4 31B IT for?
Self-hosted Google open weights for privacy-sensitive workflows that can't call cloud APIs.
What does Gemma 4 31B IT cost?
Gemma 4 31B IT lists — input and — output per 1M tokens. Pricing last checked 2026-06-09.
How large is the Gemma 4 31B IT context window?
Gemma 4 31B IT publishes a 131K token context window.
What is the VerdictPal status of Gemma 4 31B IT?
Gemma 4 31B IT carries Solid status on the model atlas.
The Pack · Editorial newsletter
Know when the rules move.
One short email when a university changes its AI rule for theses, or a research tool changes its pricing or privacy terms. No tracking, no third-party analytics. Unsubscribe in one click.