Kimi K2 Instruct 0905 vs Hermes 2 Pro Llama 3 8B vs Apertus 70B
Hermes 2 Pro Llama 3 8B comes out ahead, 51 to 38 and 32 on our weighted score, and it is the cheaper option too.
Moonshot AIKimi K2 Instruct 0905
38/100- ECI—
- Price$0.60 / $2.50
- Context262K
- Our pick
Nous ResearchHermes 2 Pro Llama 3 8B
51/100- ECI—
- Price$0.14 / $0.14
- Context8K
Swiss AI
Apertus 70B
32/100- ECI—
- Price$0.82 / $2.42
- Context66K
Hermes 2 Pro Llama 3 8B is our pick
Hermes 2 Pro Llama 3 8B is the better all-round choice, scoring 51/100 against Kimi K2 Instruct 0905 (38) and Apertus 70B (32). It leads on price. Kimi K2 Instruct 0905 wins on context window. The score weighs price 50%, inputs & features 30%, context window 20%. None of the shared benchmarks cover every model here yet, so capability is left out of this verdict.
- CapabilityNot enough dataNo independent benchmark covers every model here yet
- Lowest priceHermes 2 Pro Llama 3 8BHermes 2 Pro Llama 3 8B $0.14 · Kimi K2 Instruct 0905 $1.07 · Apertus 70B $1.22 per 1M tokens (3:1 blend)
- Longest contextKimi K2 Instruct 0905Kimi K2 Instruct 0905 262,144 · Apertus 70B 65,536 · Hermes 2 Pro Llama 3 8B 8,192 tokens
- Widest inputsSame inputsKimi K2 Instruct 0905: Text · Hermes 2 Pro Llama 3 8B: Text · Apertus 70B: Text
- Self-hostingAll open weightsEvery model here can be downloaded and run on your own hardware
| Measure | Weight | Kimi K2 Instruct 0905 | Hermes 2 Pro Llama 3 8B | Apertus 70B |
|---|---|---|---|---|
| Price | 50% | 46 | 87 | 44 |
| Inputs & features | 30% | 25 | 25 | 25 |
| Context window | 20% | 37 | 0 | 12 |
| Overall | 100% | 38/100 | 51/100 | 32/100 |
Left out because at least one model lacks the data: capability. The remaining weights were rescaled.
Every spec in one table
Highlighted cells lead their row. Dashes mean the data is not published.
| Specification | Kimi K2 Instruct 0905Moonshot AI | Hermes 2 Pro Llama 3 8BNous Research | Apertus 70BSwiss AI |
|---|---|---|---|
| Capability | |||
| Capabilities Index (ECI) | — | — | — |
| ECI rank | — | — | — |
| Price per million tokens | |||
| Input | $0.60 | $0.14 (best) | $0.82 |
| Output | $2.50 | $0.14 (best) | $2.42 |
| Cached input | — | — | — |
| Blended (3:1) | $1.07 | $0.14 (best) | $1.22 |
| Long-context rate | Same rate | Same rate | Same rate |
| Price source | Median of 12 providers | Median of 2 providers | Median of 3 providers |
| Limits | |||
| Context window | 262,144 tokens (best) | 8,192 tokens | 65,536 tokens |
| Max output | 98,304 tokens (best) | 8,192 tokens | 8,192 tokens |
| Inputs and features | |||
| Text | Yes | Yes | Yes |
| Images | No | No | No |
| PDFs | No | No | No |
| Audio | No | No | No |
| Video | No | No | No |
| Reasoning | No | No | No |
| Tool calling | Yes | Yes | Yes |
| Structured output | No | No | No |
| Availability | |||
| Weights | OpenModified MIT | OpenLlama 3 Community License | OpenApache-2.0 |
| API model ID | — | — | — |
| API providers | 13 (best) | 2 | 3 |
| Released | Sep 5, 2025 | Apr 30, 2024 | Sep 2, 2025 |
| Knowledge cutoff | — | — | Sep 2025 |
What would a month cost?
Enter your expected volume in millions of tokens. List prices only; caching and batch discounts would lower these.
- Kimi K2 Instruct 0905$11.00
- Hermes 2 Pro Llama 3 8B$1.68
- Apertus 70B$13.04
Which should you choose?
Which is better: Kimi K2 Instruct 0905, Hermes 2 Pro Llama 3 8B or Apertus 70B?
Hermes 2 Pro Llama 3 8B is the better all-round choice, scoring 51/100 against Kimi K2 Instruct 0905 (38) and Apertus 70B (32). It leads on price. Kimi K2 Instruct 0905 wins on context window. The score weighs price 50%, inputs & features 30%, context window 20%. None of the shared benchmarks cover every model here yet, so capability is left out of this verdict.
Which is cheaper, Kimi K2 Instruct 0905, Hermes 2 Pro Llama 3 8B or Apertus 70B?
Hermes 2 Pro Llama 3 8B is cheaper at $0.14 input / $0.14 output per million tokens (median across 2 API providers). Kimi K2 Instruct 0905 costs $0.60 input / $2.50 output per million tokens (median across 12 API providers); Apertus 70B costs $0.82 input / $2.42 output per million tokens (median across 3 API providers). At a typical mix of three input tokens to one output token, that is $0.14 per million tokens for Hermes 2 Pro Llama 3 8B versus $1.07 for Kimi K2 Instruct 0905 (7.7× as much) and $1.22 for Apertus 70B (8.7× as much).
Which scores higher on benchmarks?
There is no independent benchmark that covers all three models yet. Kimi K2 Instruct 0905 has not been scored yet, Hermes 2 Pro Llama 3 8B has not been scored yet and Apertus 70B has not been scored yet.
Which is better for coding?
There are no published SWE-bench Verified results for Kimi K2 Instruct 0905, Hermes 2 Pro Llama 3 8B and Apertus 70B yet, so there is no like-for-like coding score. Test both on a sample of your own tasks. All three support tool calling for agent workflows.
Which has the bigger context window?
Kimi K2 Instruct 0905 has the largest context window at 262,144 tokens, against 65,536 for Apertus 70B and 8,192 for Hermes 2 Pro Llama 3 8B. Maximum output per response: Kimi K2 Instruct 0905 up to 98,304, Hermes 2 Pro Llama 3 8B up to 8,192, Apertus 70B up to 8,192 tokens.
Which can read images, PDFs, audio or video?
Kimi K2 Instruct 0905 accepts text; Hermes 2 Pro Llama 3 8B accepts text; Apertus 70B accepts text. They handle the same number of input types.
Are any of these open source?
Yes, all three publish their weights (Modified MIT, Llama 3 Community License and Apache-2.0), so you can self-host them.
Which is newer?
Kimi K2 Instruct 0905 is the newest, released Sep 5, 2025. Apertus 70B came out Sep 2, 2025; Hermes 2 Pro Llama 3 8B came out Apr 30, 2024. Knowledge cutoff: Apertus 70B Sep 2025.
How do you decide the winner?
Each model gets a 0–100 score on capability (50%, independent benchmark results); price (25%, blended price per million tokens (3 input : 1 output), log scale); inputs & features (15%, image, PDF, audio and video input, tool calling, structured output and reasoning); context window (10%, maximum tokens per request, log scale). Dimensions missing for any model are dropped and the remaining weights rescaled, so every model is judged on the same evidence. Specs and prices come from public model listings and the labs’ own API pages; capability scores come from independent benchmark runs. Data updated Nov 2, 2026.
