of MiMo-V2.6-Pro's tool calls fail on NovitaAI (fp8).
Not all models are served equally by all providers. On a typical model, the worst provider breaks tool calls 19× as often as the best. Find the best one for your agents.
| # | Provider | Broken tool calls | Tool calls a day | $ in / out per M | Tokens/s | Uptime |
|---|---|---|---|---|---|---|
| 1 | DeepInfrafp8# | 1.48% | 244k | $0.39 / $0.78 | 35 | 99.8% |
| 2 | GMICloudbf16# | 3.77% | 227k | $0.37 / $0.74 | 27 | 99.7% |
| 3 | Xiaomifp8# | 4.59% | 656k | $0.43 / $0.87 | 26 | 99.6% |
| 4 | NovitaAIfp8# | 5.47% | 343k | $0.43 / $0.87 | 30 | 99.5% |
Also ignored, under 1,000 tool calls a day: StreamLake 5.01%.
{
"provider": {
"order": [
"deepinfra/fp8"
],
"ignore": [
"gmicloud/bf16",
"xiaomi/fp8",
"streamlake",
"novita/fp8"
]
}
}Point your agent at llms.txt
And which provider to use.
Releasing a model? Test it before launch.
of MiMo-V2.6-Pro's tool calls fail on NovitaAI (fp8).
as often on NovitaAI (fp8) as on DeepInfra (fp8), its best provider.
broken tool calls a day across its 4 providers.