# Gemma 4 12B

Google · Open Weight · rank 157 · bench-align-v5

> Every figure below is reproduced as its upstream source published it: nothing is modelled, estimated, interpolated or converted. `Unavailable` means no source published the value — it is never a zero. Each value carries its evidence state and the date it was observed.

Page: https://modelscale.dev/models/gemma-4-12b  
JSON: https://modelscale.dev/api/model/gemma-4-12b

## Facts

| Field | Value | Evidence |
| --- | --- | --- |
| Canonical id | `gemma-4-12b` | — |
| Overall score | 43.24 | Observed 2026-09-22 · source benchlm:models |
| Context window | 256K tokens | Observed 2026-09-22 · source benchlm:models |
| Release date | 2026-06-03 | Observed 2026-09-22 · source benchlm:models |
| Access type | Open Weight | — |
| Blended $/1M (75% input / 25% output) | Unavailable | Derived from the input and output rates below |

## Capability evidence

Seven axes from the ranking source. An axis the source did not score is unavailable, not zero.

| Axis | Score | Evidence |
| --- | --- | --- |
| Agentic | Unavailable | Unavailable · source benchlm:models |
| Coding | Unavailable | Unavailable · source benchlm:models |
| Knowledge | 53.5 | Observed 2026-09-22 · source benchlm:models |
| Reasoning | 33.7 | Observed 2026-09-22 · source benchlm:models |
| Multimodal & Grounded | 26.2 | Observed 2026-09-22 · source benchlm:models |
| Instruction Following | 88.7 | Observed 2026-09-22 · source benchlm:models |
| Math | 53 | Observed 2026-09-22 · source benchlm:models |

## Runtime service evidence

Measured values with the date they were observed. Nothing is inferred from a sibling model or a provider claim. Regional or per-endpoint measurements appear only when the API supplies them; none are modelled.

| Measurement | Value | Observed | Last good | Evidence |
| --- | --- | --- | --- | --- |
| Time to first token | Unavailable | Unavailable | Unavailable | Unavailable · source benchlm:speed |
| Throughput | Unavailable | Unavailable | Unavailable | Unavailable · source benchlm:speed |

## Endpoint and price matrix

Every published price component, including cache reads and writes.

| Component | USD | Evidence |
| --- | --- | --- |
| Input / 1M tokens | Unavailable | Unavailable · source benchlm:pricing |
| Output / 1M tokens | Unavailable | Unavailable · source benchlm:pricing |
| Cache read / 1M tokens | Unavailable | Unavailable · source benchlm:pricing |
| Cache write / 1M tokens | Unavailable | Unavailable · source openrouter:pricing |
| Blended / 1M (75% input / 25% output) | Unavailable | Derived — from the input and output rates above; it has no source record of its own |
| Cost per successful task (LiveBench) | Unavailable | Unavailable · source livebench:table |

## Workload-aware monthly cost example

10 conversations per day × 8 messages × 22 active days, 1200 input and 400 output tokens per message, no cache. Derived here from the published rates above by this site's own calculator — not a figure any source published.

| Field | Value |
| --- | --- |
| Modelled monthly cost | Unavailable |
| Modelled tokens | Unavailable |
| Reason | The applicable input rate is unavailable. |

## Benchmark record

26 matched benchmark rows with their published value, unit, and provenance.

### Knowledge

| Benchmark | Value | Tasks | Difficulty | Provenance |
| --- | --- | --- | --- | --- |
| GPQA (Graduate-Level Google-Proof Q&A) | 78.8 | 448 questions | Graduate level | [GPQA: A Graduate-Level Google-Proof Q&A Benchmark](https://arxiv.org/abs/2311.12022) |
| GPQA-D (GPQA Diamond) | 78.8 | Graduate-level science questions | Graduate level | [Trinity-Large-Thinking: Scaling an Open Source Frontier Agent](https://www.arcee.ai/blog/trinity-large-thinking) |
| MMLU-Pro (Massive Multitask Language Understanding Professional) | 77.2 | Multiple subjects | Professional level | [MMLU-Pro: A More Robust and Challenging Multi-Task Language Understanding Benchmark](https://arxiv.org/abs/2406.01574) |
| Artificial Analysis Intelligence Index | 14.2 | Cross-benchmark intelligence index | Display-only external reference | [Artificial Analysis](https://artificialanalysis.ai/) |
| AA-GPQA Diamond (Artificial Analysis GPQA Diamond) | 75.3 | Graduate-level science questions | Graduate-level science reasoning | [Artificial Analysis GPQA Diamond Benchmark Leaderboard](https://artificialanalysis.ai/evaluations/gpqa-diamond) |
| AA-HLE (Artificial Analysis Humanity's Last Exam) | 15.7 | Expert-level questions | Frontier expert reasoning | [Artificial Analysis Humanity's Last Exam Benchmark Leaderboard](https://artificialanalysis.ai/evaluations/hle) |
| AA-Omniscience Index (Artificial Analysis Omniscience Index) | -52.7 | Knowledge questions | Broad factual knowledge | [AA-Omniscience: Knowledge and Hallucination Benchmark](https://artificialanalysis.ai/evaluations/omniscience) |
| AA-Omniscience Accuracy (Artificial Analysis Omniscience Accuracy) | 15.6 | Knowledge questions | Broad knowledge | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/grok-4-3) |
| AA-Omniscience Hallucination Rate (Artificial Analysis Omniscience Hallucination Rate) | 81.0 | Knowledge questions | Factuality | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/grok-4-3) |
| HLE w/o tools (Humanity's Last Exam without tools) | 5.2 | Expert-level questions | Frontier expert level | [Introducing GPT-5.4 mini and nano](https://openai.com/index/introducing-gpt-5-4-mini-and-nano/) |
| MMMLU | 83.4 | Multilingual academic QA | Broad multilingual knowledge | [MMMLU](https://huggingface.co/datasets/openai/MMMLU) |

### Coding

| Benchmark | Value | Tasks | Difficulty | Provenance |
| --- | --- | --- | --- | --- |
| LiveCodeBench v6 | 72.0 | Fresh programming problems | Competitive programming level | [LiveCodeBench official repository and release documentation](https://github.com/LiveCodeBench/LiveCodeBench) |
| AA Coding Index (Artificial Analysis Coding Index) | 31.0 | Cross-benchmark coding index | Display-only external reference | [Artificial Analysis model leaderboards](https://artificialanalysis.ai/leaderboards/models) |

### Mathematics

| Benchmark | Value | Tasks | Difficulty | Provenance |
| --- | --- | --- | --- | --- |
| AIME26 (AIME 2026) | 77.5 | Competition math problems | Olympiad-style mathematics | [Qwen3.6 launch benchmarks](https://qwen.ai/blog?id=qwen3.6) |

### Reasoning

| Benchmark | Value | Tasks | Difficulty | Provenance |
| --- | --- | --- | --- | --- |
| BBH (BIG-Bench Hard) | 53 | 23 tasks | Advanced reasoning | [Challenging BIG-Bench Tasks and Whether Chain-of-Thought Can Solve Them](https://arxiv.org/abs/2210.09261) |
| MRCRv2 | 43.4 | Long-context retrieval | Hard long-context | [Introducing GPT-5.2 and GPT-5.2 Pro](https://openai.com/index/introducing-gpt-5-2/) |
| AA-LCR (Artificial Analysis Long Context Reasoning) | 63.7 | Long-context reasoning tasks | Long-context reasoning | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/grok-4-3) |
| CritPt (Critical Physics Tasks) | 0.0 | Research-level physics questions | Research-level physics reasoning | [CritPt Benchmark Leaderboard](https://artificialanalysis.ai/evaluations/critpt) |

### Instruction Following

| Benchmark | Value | Tasks | Difficulty | Provenance |
| --- | --- | --- | --- | --- |
| AA-IFBench (Artificial Analysis IFBench) | 73.5 | Verifiable instruction constraints | Instruction precision | [Artificial Analysis IFBench Benchmark Leaderboard](https://artificialanalysis.ai/evaluations/ifbench) |

### Agentic

| Benchmark | Value | Tasks | Difficulty | Provenance |
| --- | --- | --- | --- | --- |
| GDPval-AA | 591 | Agentic real-world work tasks | Professional agentic workflows | [DeepSeek-V4 Technical Report](https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro/blob/main/DeepSeek_V4.pdf) |
| GDPval-AA (GDPval-AA normalized) | 0.0 | Economically valuable tasks | Professional agentic workflows | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/grok-4-3) |
| τ²-bench results (τ²-Bench Tool-Agent-User Evaluation) | 36.3 | Airline, retail, and telecom customer-service task sets | Dual-control customer-service workflows | [τ²-Bench: Evaluating Conversational Agents in a Dual-Control Environment](https://arxiv.org/abs/2506.07982) |

### Multimodal & Grounded

| Benchmark | Value | Tasks | Difficulty | Provenance |
| --- | --- | --- | --- | --- |
| MMMU-Pro (Massive Multi-discipline Multimodal Understanding Pro) | 69.1 | Multimodal academic reasoning | Frontier multimodal | [MMMU-Pro: A More Robust Multi-discipline Multimodal Understanding Benchmark](https://arxiv.org/abs/2409.02813) |
| AA-MMMU-Pro (Artificial Analysis MMMU-Pro) | 69.7 | Multimodal academic reasoning | Frontier multimodal | [Artificial Analysis MMMU-Pro Benchmark Leaderboard](https://artificialanalysis.ai/evaluations/mmmu-pro) |
| MathVision | 79.7 | Visually grounded math problems | Advanced multimodal mathematics | [Qwen3.6 launch benchmarks](https://qwen.ai/blog?id=qwen3.6) |
| MedXpertQA (MM) (MedXpertQA Multimodal) | 48.7 | 2,000 multimodal medical questions | Clinical multimodal reasoning | [Muse Spark Eval Methodology](https://ai.meta.com/static-resource/muse-spark-eval-methodology) |

## Lifecycle and limitations log

Lifecycle events the source associates with this model.

No lifecycle event references this model. That is not evidence the model has no lifecycle plan — only that this source published none.

## What this profile does not claim

Values are reproduced exactly as their sources published them, in the units those sources declared; none are converted, interpolated, or averaged across providers. Any field marked unavailable was attempted and not returned. Last attempted fetch for this model's score: 2026-09-22 10:17 UTC.