Swiss AIApertus 8B Instruct
Apertus 8B Instruct is a Swiss AI open-weights model, released 2 Sept 2025. Its best configuration ranks #731 of 785 on the BenchLeader Index at 37.4 ±4.7, in the lower half. The ± is the point: 196 other configurations score within that range, so they and this one cannot be told apart on quality alone — price and speed are what separate them. It scores highest in reasoning (41) and lowest in long context (23). At $0.125 per million tokens blended it is among the cheapest fifth of ranked models. Output speed of 131 tokens per second puts it faster than most, with a first answer in 3.2 s. Last measured 2 Sept 2025.stale: no new result in six months
- Blended price
- $0.125/M
- $0.100 in · $0.200 out
- Output speed
- 131 tok/s
- measured by Artificial Analysis
- First answer
- 3.22 s
- first token 3.22 s
- Context
- 66k
- Full answer
- 7.04 s
- median, reasoning included
- Cost per run
- –
- Released
- 2 Sept 2025
- Overall index37
- Reasoning41
- Agents & tools34
- Knowledge27
- Instruction following27
- Long context23
- Composite35
How its index has moved
9 Sept 2026 to 21 Oct 2026The index is recomputed from scratch every day, so a line moves when a new benchmark result lands, when a publisher revises a score, or when the models it is normalised against change. Early movement usually means the score is still settling.
Benchmark results
See all LLM benchmarks. One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.
Metadata
| Benchmark | not stated | Source |
|---|---|---|
| AA Openness Indexnot in index | 88.9#1 | Artificial Analysis |
Reasoning
| Benchmark | not stated | Source |
|---|---|---|
| GPQA Diamond (AA)not in index | 25.6%#544 | Artificial Analysis |
| Humanity's Last Exam (AA)not in index | 5.0%#459 | Artificial Analysis |
| CritPt | 0.0%#315 | Artificial Analysis |
Agents & tools
| Benchmark | not stated | Source |
|---|---|---|
| Terminal-Bench Hard | 0.0%#362 | Artificial Analysis |
| τ²-Bench Telecom (AA)not in index | 11.4%#376 | Artificial Analysis |
Knowledge
| Benchmark | not stated | Source |
|---|---|---|
| AA-Omniscience | -77.2#533 | Artificial Analysis |
| AA-Omniscience: accuracynot in index | 9.7%#499 | Artificial Analysis |
| AA-Omniscience: non-hallucinationnot in index | 3.9%#537 | Artificial Analysis |
Instruction following
| Benchmark | not stated | Source |
|---|---|---|
| IFBench | 22.4%#405 | Artificial Analysis |
Long context
| Benchmark | not stated | Source |
|---|---|---|
| AA-LCR | 0.0%#493 | Artificial Analysis |
Composite
| Benchmark | not stated | Source |
|---|---|---|
| AA Intelligence Index v4.3.2 | 4.8#618 | Artificial Analysis |
What a task costs
Estimates from list price, output speed and time to first answer for the best configuration. “With caching” assumes three-quarters of the input is served from the prompt cache. Reasoning tokens are not modelled.
| Workload | Tokens in / out | Cost | With caching | Time |
|---|---|---|---|---|
| Chat reply | 400 / 300 | $0.0001 | – | 5.5 s |
| Summarize a 30-page report | 12,000 / 600 | $0.0013 | – | 7.8 s |
| Code edit | 6,000 / 1,500 | $0.0009 | – | 14.7 s |
| Agentic coding session | 60,000 / 4,000 | $0.0068 | – | 33.8 s |
| Structured extraction | 2,000 / 200 | $0.0002 | – | 4.7 s |
More from Swiss AI
See also
Questions people ask
- What are Apertus 8B Instruct's benchmark scores?
- As of 25 Oct 2026, Apertus 8B Instruct scores 37.4 ±4.7 on the BenchLeader Index, strongest in reasoning (41) and weakest in long context (23). That rests on 6 independent benchmark results, and every score on this page links to its publisher.
- How does Apertus 8B Instruct rank?
- Apertus 8B Instruct is #731 of 785 ranked configurations on 25 Oct 2026. 196 other configurations score within its margin, so the exact position matters less than the band. The leader is Claude Opus 5.5 (max effort) at 72.6.
- How much does Apertus 8B Instruct cost per 1M tokens?
- As of 25 Oct 2026, Apertus 8B Instruct costs $0.100 per 1M input tokens and $0.200 per 1M output tokens, $0.125 blended (three input tokens per output token).
- When was Apertus 8B Instruct released?
- Swiss AI released Apertus 8B Instruct on 2 Sept 2025.
Data as of 25 Oct 2026. Compare with another model.
Cite as: BenchLeader, “Apertus 8B Instruct: benchmarks, pricing, speed and rank”, data as of 25 Oct 2026, https://www.benchleader.com/models/apertus-8b.