Inference provider
DekaLLM
8 models served. Speed, latency and price below are medians across this provider's catalogue, not per-endpoint measurements.
Models served
8
Median output speed
112 t/s
Median latency
283ms
Median price / 1M
$0.161
Catalogue
Strongest and fastest on this provider
What the catalogue looks like at the top end, on the two axes that usually decide the choice.
Highest Intelligence Index
Among models this provider serves
- Qwen3.8 27BQwen33.7
- Qwen3.6 35B A3BQwen18.2
- gpt-oss-120bOpenAI11.6
- gpt-oss-20bOpenAI9.0
View as table
| Model | Intelligence |
|---|---|
| Qwen3.8 27B (Qwen) | 33.7 |
| Qwen3.6 35B A3B (Qwen) | 18.2 |
| gpt-oss-120b (OpenAI) | 11.6 |
| gpt-oss-20b (OpenAI) | 9.0 |
Fastest output
Median tokens per second
- gpt-oss-120bOpenAI641 t/s
- gpt-oss-20bOpenAI410 t/s
- Qwen3.8 27BQwen409 t/s
- Qwen3.6 35B A3BQwen128 t/s
- Mistral NemoMistral AI96 t/s
- Gemma 4 26B A4B Google95 t/s
- Nemotron 3 SuperNVIDIA74 t/s
- 74 t/s
View as table
| Model | Tokens/s |
|---|---|
| gpt-oss-120b (OpenAI) | 641 |
| gpt-oss-20b (OpenAI) | 410 |
| Qwen3.8 27B (Qwen) | 409 |
| Qwen3.6 35B A3B (Qwen) | 128 |
| Mistral Nemo (Mistral AI) | 96 |
| Gemma 4 26B A4B (Google) | 95 |
| Nemotron 3 Super (NVIDIA) | 74 |
| Qwen3 30B A3B Instruct 2507 (Qwen) | 74 |
Full catalogue
All 8 models on DekaLLM
Filter by creator
8 of 8 models| Qwen3.8 27B Qwenopen | 33.7 | $0.19 |
| Qwen3.6 35B A3B Qwenopen | 18.2 | $0.09 |
| gpt-oss-120b OpenAIopen | 11.6 | $0.06 |
| gpt-oss-20b OpenAIopen | 9.0 | $0.008 |
| Gemma 4 26B A4B Googleopen | — | $0.03 |
| Nemotron 3 Super NVIDIAopen | — | $0.04 |
| Qwen3 30B A3B Instruct 2507 Qwenopen | — | $0.01 |
| Mistral Nemo Mistral AIopen | — | $0.002 |