of GLM 5.1's tool calls fail on StreamLake (fp8).
Updated October 29, 2026
Tool-call index
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 | Friendli# | 0.00% | 1.3k | $1.40 / $4.40 | 161 | 99.9% |
| 2 | Alibaba Cloud Int.fp8# | 0.05% | 1.1k | $1.33 / $4.18 | 95 | 96.9% |
| 3 | Chutesfp8# | 0.87% | 2.4k | $0.90 / $2.90 | 29 | 98.8% |
| 4 | Z.aifp8# | 1.45% | 12k | $1.40 / $4.40 | 39 | 99.8% |
| 5 | SiliconFlowfp8# | 2.43% | 7.0k | $1.19 / $3.74 | 83 | 99.9% |
| 6 | Baidu Qianfanfp8# | 4.44% | 2.9k | $1.40 / $4.40 | 66 | 99.9% |
| 7 | StreamLakefp8# | 6.96% | 2.9k | $0.97 / $3.04 | 30 | 99.8% |
Also ignored, under 1,000 tool calls a day: Phala 3.92%.
Recommended OpenRouter settings for GLM 5.1
{
"provider": {
"order": [
"friendli",
"alibaba/fp8",
"chutes/fp8"
],
"ignore": [
"phala",
"baidu/fp8",
"streamlake/fp8"
]
}
}Point your agent at llms.txt
Get updates when its successor drops
And which provider to use.
For model makers
Broken tool calls carry your model's name
Releasing a model? Test it before launch.
as often on StreamLake (fp8) as on Friendli, its best provider.
broken tool calls a day across its 7 providers.