Network
Batch AI providers
Compare model coverage, observed activity, and privacy posture across providers.
6
Providers
344
Models
491K
Executions
99.3%
Avg success
OP
OpenAI Batch
lane · openaiActive
Native OpenAI Batch API lane. 24h SLA with 10% public-provider discount on quotes.
Models
38
Executions
184K
Success
99.6%
chat
embeddings
responses
AN
Anthropic
lane · anthropicActive
Claude 3 / 3.5 batch lanes with strict privacy posture. 10% public-provider discount.
Models
11
Executions
122K
Success
99.8%
chat
tool-use
OP
OpenRouter
lane · openrouterActive
Edge
Aggregated lane across 200+ models. Useful as a price-aware fallback.
Models
214
Executions
76K
Success
99.1%
chat
completions
MI
Mistral
lane · mistralActive
EU-hosted batch lane for Mistral and Codestral families with residency guarantees.
Models
9
Executions
35K
Success
99.4%
chat
embeddings
TO
Together AI
lane · together-aiActive
Open-weight model lane — Llama, Qwen, DeepSeek — billed at provider rate.
Models
58
Executions
51K
Success
99.2%
chat
embeddings
rerank
GR
Groq
lane · groqActive
Edge
Low-latency LPU lane used when batches need a fast retry leg.
Models
14
Executions
23K
Success
98.9%
chat
Run a batch-capable lane?
Apply to join the routing pool — telemetry and rankings publish automatically.