GLM-5 is Z.ai’s flagship open-source foundation model engineered for complex systems design and long-horizon agent workflows. Built for expert developers, it delivers production-grade performance on large-scale programming tasks, rivaling leading...
Providers
Reference endpoint rows are source-labelled and kept separate from CLSSAI route eligibility. Temporary OpenRouter rows do not claim that CLSSAI routed a request through that Provider.
Temporary reference data from the OpenRouter public endpoints API. It will be replaced by CLSSAI-native endpoint eligibility and telemetry.
| Provider | Input $/1M | Output $/1M | Latency | Throughput | Uptime |
|---|---|---|---|---|---|
| SStreamLakeFP8 | $0.6 | $1.92 | -- | -- | 99.85% |
| GGMICloudFP8 | $0.6 | $1.92 | -- | -- | 95.6% |
| DDeepInfraFP4 | $0.6 | $2.08 | -- | -- | 100.00% |
| BBaiduFP8 | $0.7 | $2.24 | -- | -- | 99.83% |
| DDigitalOceanStandard | $0.75 | $2.4 | -- | -- | 98.9% |
| SSiliconFlowFP8 | $0.95 | $2.55 | -- | -- | 99.48% |
| AAtlasCloudFP8 | $0.95 | $3.15 | -- | -- | 100.00% |
| AAmazon BedrockStandard | $1 | $3.2 | -- | -- | 87.9% |
| NNovitaFP8 | $1 | $3.2 | -- | -- | 100.00% |
| ZZ.AIFP8 | $1 | $3.2 | -- | -- | 99.05% |
| PParasailFP8 | $1 | $3.2 | -- | -- | 98.8% |
| VVeniceFP8 | $1 | $3.2 | -- | -- | 99.24% |
| PPhalaStandard | $1.2 | $3.5 | -- | -- | -- |
Effective Pricing
Effective price reflects what requests actually cost after caching and endpoint selection. No aggregate ledger is connected to this view yet.
Weighted Avg Input Price
--
Verified telemetry pending
Weighted Avg Output Price
--
Verified telemetry pending
Token-weighted provider price, cache-hit rate, and one-day token share require verified usage aggregation.
No verified historical input-price series is connected.
No verified historical output-price series is connected.
Performance
Throughput measures output speed; latency and time-to-first-token measure how long users wait. Only sourced telemetry belongs in these charts.
Throughput
--
Verified telemetry pending
Latency
--
Verified telemetry pending
Verified provider time-series data is not connected.
Verified provider time-series data is not connected.
Verified provider time-series data is not connected.
Verified provider time-series data is not connected.
Verified provider time-series data is not connected.
Verified provider time-series data is not connected.
Verified provider time-series data is not connected.
Uptime
Uptime is the percentage of successful requests. Failover and finish-reason claims require verified request outcomes.
Avg. Provider Uptime · 3d
--
Verified telemetry pending
No verified uptime history is connected.
No verified completion outcome series is connected.
Benchmarks
Standardized evaluations must identify their evaluator, methodology, score scale, and comparison set.
No sourced benchmark record was supplied to this page, so no score, percentile, rank, or Elo is shown.
Apps
Public applications sending verified traffic can reveal real workloads without exposing private request content.
No verified public application traffic is connected.
No verified token-volume series is connected.
Activity
Token volume and request traffic over time appear here only after aggregate, source-labelled activity is available.
Prompt, completion, reasoning, and request series are unavailable until verified aggregation is connected.
Frequently asked questions
Answers below are derived only from the endpoint view passed to this page.
Which providers serve Z Ai: Z.ai: GLM 5?
StreamLake, GMICloud, DeepInfra, Baidu, DigitalOcean, SiliconFlow, AtlasCloud, Amazon Bedrock, Novita, Z.AI, Parasail, Venice, Phala are present in the source-labelled endpoint view supplied to this page.
What context length is available for Z Ai: Z.ai: GLM 5?
The largest context length in the supplied endpoint view is 204.8K tokens.
Which API parameters does Z Ai: Z.ai: GLM 5 support?
The supplied endpoint metadata lists 19 distinct supported parameters. Open an endpoint row to inspect them.