Swiss AI Initiative · Apertus
Apertus 70B Instruct
Open 70B instruct model from the Swiss AI Initiative, served first-party by the initiative.
- AA Intelligence
- 5.1#139 of 142
- Input / output per 1M
- $0.82/ $2.92
- Output speed
- —
- Context window
- 66K tokens
Composite indexes
- Artificial Analysis Intelligence Index#139 of 1425
Leader: Claude Opus 5.5 at 58
Agents and tool use
- τ²-Bench#78 of 8812.9%
Leader: GLM-5.2 at 99.1%
Reasoning and knowledge
- GPQA Diamond#142 of 14227.2%
Leader: GPT-6 Astra at 96.1%
- Humanity’s Last Exam#118 of 1475.5%
Leader: Claude Opus 5.5 at 61.4%
- CritPt#139 of 1390%
Leader: GPT-5.6 Sol at 32.3%
Vision and long context
- AA Long Context Reasoning#139 of 1390%
Leader: Kimi K3 at 88.7%
Other published results
| Result | Score | Reported by | Source |
|---|---|---|---|
| Terminal-Bench Hard (AA) · non-reasoning | 0% | independent | Artificial Analysis: Apertus 70B Instruct ↗ |
| IFBench · non-reasoning | 25.9% | independent | Artificial Analysis: Apertus 70B Instruct ↗ |
| AA-Omniscience Index · non-reasoning | -61.5 index | independent | Artificial Analysis: Apertus 70B Instruct ↗ |