of MiMo-V2.6-Flash's tool calls fail on io.net (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 | Venicefp8# | 1.41% | 32k | $0.17 / $0.35 | 60 | 65.5% |
| 2 | Xiaomifp8# | 1.71% | 3.8M | $0.14 / $0.28 | 43 | 100.0% |
| 3 | DeepInfrafp8# | 2.12% | 178k | $0.14 / $0.28 | 68 | 76.2% |
| 4 | Cohere# | 2.70% | 290k | $0.14 / $0.28 | 88 | 99.7% |
| 5 | GMICloudbf16# | 2.79% | 29k | $0.12 / $0.24 | 33 | 96.3% |
| 6 | NovitaAIfp8# | 2.82% | 76k | $0.14 / $0.28 | 17 | 98.5% |
| 7 | Makorafp8# | 3.92% | 149k | $0.13 / $0.28 | 65 | 99.2% |
| 8 | Relacemxfp4# | 4.17% | 211k | $0.13 / $0.28 | 101 | 100.0% |
| 9 | io.netfp8# | 4.73% | 76k | $0.10 / $0.28 | 65 | 99.3% |
{
"provider": {
"order": [
"venice/fp8",
"xiaomi/fp8",
"deepinfra/fp8"
],
"ignore": [
"makora/fp8",
"relace/mxfp4",
"io-net/fp8"
]
}
}Point your agent at llms.txt
And which provider to use.
Releasing a model? Test it before launch.
of MiMo-V2.6-Flash's tool calls fail on io.net (fp8).
as often on io.net (fp8) as on Venice (fp8), its best provider.
broken tool calls a day across its 9 providers.