Drafted with AI tools and edited by a human. The figures and conclusions were checked by a GPUniq editor.
Fast LLM API Benchmarks
GPT-6 Luna offers the lowest input cost at $0.08 per 1M tokens ($7.59 RUB). MiMo V2.5 delivers 80.3 tokens per second alongside an output cost of $0.357 per 1M tokens ($33.86 RUB). Qwen3.8 Flash hits 16.7 tokens per second with a 1,000K token context window. Choosing the best fast model depends on balancing input-heavy workload costs against required generation speeds.
Which model is the cheapest API?
GPT-6 Luna leads input pricing at $0.08 per 1M tokens ($7.59 RUB). Qwen3.8 Flash sits in the middle at $0.135 per 1M tokens ($12.80 RUB). MiMo V2.5 has the highest input cost at $0.178 per 1M tokens ($16.93 RUB).
| Model | Input Price, $/1M tokens | Input Price, RUB/1M tokens |
|---|---|---|
| GPT-6 Luna | 0.08 | 7.59 |
| Qwen3.8 Flash | 0.135 | 12.80 |
| MiMo V2.5 | 0.178 | 16.93 |
How do output token costs compare?
MiMo V2.5 provides the cheapest generation at $0.357 per 1M tokens ($33.86 RUB). GPT-6 Luna charges $0.40 per 1M tokens ($37.94 RUB) for output generation. Qwen3.8 Flash is the most expensive for outputs at $0.423 per 1M tokens ($40.12 RUB).
| Model | Output Price, $/1M tokens | Output Price, RUB/1M tokens |
|---|---|---|
| MiMo V2.5 | 0.357 | 33.86 |
| GPT-6 Luna | 0.40 | 37.94 |
| Qwen3.8 Flash | 0.423 | 40.12 |
Which API delivers the fastest speed?
GPT-6 Luna achieves top throughput at 81.3 tokens per second. MiMo V2.5 follows closely with 80.3 tokens per second. Qwen3.8 Flash lags far behind at only 16.7 tokens per second.
| Model | Speed, tokens/sec | Vision Support |
|---|---|---|
| GPT-6 Luna | 81.3 | Yes |
| MiMo V2.5 | 80.3 | No |
| Qwen3.8 Flash | 16.7 | No |
What are the context window limits?
MiMo V2.5 supports a 1,050K token context window. GPT-6 Luna matches this capacity with a 1,050K token limit. Qwen3.8 Flash offers a 1,000K token context window.
| Model | Context Window, K tokens |
|---|---|
| MiMo V2.5 | 1050 |
| GPT-6 Luna | 1050 |
| Qwen3.8 Flash | 1000 |
What API discounts are available?
MiMo V2.5 features a 30% discount off standard list prices. GPT-6 Luna includes a 20.0% promotional API discount. Qwen3.8 Flash offers a 10.0% discount on gateway traffic.
| Model | Gateway Discount, % |
|---|---|
| MiMo V2.5 | 30 |
| GPT-6 Luna | 20.0 |
| Qwen3.8 Flash | 10.0 |
Related on the GPUniq blog
Methodology
Prices are GPUniq's public retail rates per million tokens, read from the live model catalog at the moment of the snapshot. Ruble figures use the rate GPUniq applies to ruble top-ups at that moment. The discount is against the model vendor's own list price. Speed is output tokens per second measured on the GPUniq gateway for real requests, as a moving average; models with too little traffic to measure are left out rather than estimated.
Sources
- 1.GPUniq model catalog and prices — GPUniq (accessed )
Want cheap GPUs for your next project?
Browse live GPU prices and rent the right card in seconds — H100, A100, RTX 4090, and 50+ more models.
