DekaLLM — pricing & performance
Models served
7
of 1,120 tracked models
Mean output speed
51
tokens per second, averaged over
7 measured endpoints
7 measured endpoints
Mean latency
3.2s
time to first token, averaged over
7 measured endpoints
7 measured endpoints
Cheapest blended
$0.021
per 1M tokens at a 3:1
input:output blend
input:output blend
Output speed on DekaLLM
Output tokens per second for all 7 models on DekaLLM with published throughput telemetry, the most recent figure OpenRouter publishes for each host. Higher is better. We do not run these measurements ourselves.
All 7 of 7 models⇩
The Known Good
7 models on DekaLLM
Full leaderboard →
| Model | Creator | Input $/M | Output $/M | Blended | Speed | TTFT | Context |
|---|---|---|---|---|---|---|---|
Google: Gemma 4 26B A4B
|
$0.060 | $0.33 | $0.13 | 70 | 0.7s | 262k | |
Mistral: Mistral Nemo
|
Mistral | $0.018 | $0.030 | $0.021 | 10 | 0.9s | 131k |
NVIDIA: Nemotron 3 Super
|
NVIDIA | $0.080 | $0.45 | $0.17 | 3 | 16.9s | 262k |
OpenAI: gpt-oss-120b
|
OpenAI | $0.030 | $0.18 | $0.068 | 21 | 1.1s | 131k |
|
OpenAI: gpt-oss-20b
|
OpenAI | $0.029 | $0.14 | $0.057 | 14 | 1.7s | 131k |
Qwen: Qwen3 30B A3B Instruct 2507
|
Alibaba | $0.090 | $0.30 | $0.14 | 64 | 0.4s | 262k |
|
Qwen: Qwen3.6 35B A3B
|
Alibaba | $0.13 | $1.00 | $0.35 | 174 | 0.6s | 262k |
The Known Good
Source. Prices, endpoint availability and throughput for DekaLLM are ingested from OpenRouter;
model metadata is enriched from models.dev. A blank cell means the figure was not published for that endpoint —
it never means zero. Blended price is (3 × input + output) ÷ 4, the same 3:1 blend used everywhere on this
site; see methodology. Quality scores are not provider-specific:
a model scores the same wherever it is hosted, so quality lives on the
model pages.