Provider guide
Where to run Claude Haiku 4.5
10 live listings tracked.
FASTEST MEASURED
195 tok/s
measured throughput · $1.10 input / $5.50 output per million tokens
OpenRouter$1.00 in/1M$5.00 out/1M—200K—30 min agoDeepInfra$1.00 in/1M$5.00 out/1M—200K—6 hours agoMicrosoft Azure AI$1.00 in/1M$5.00 out/1M65 tok/s200K—30 min agoGoogle Vertex AIzero-retention$1.00 in/1M$5.00 out/1M85 tok/s200K—30 min agoAnthropicCHEAPEST$1.00 in/1M$5.00 out/1M72 tok/s200K—30 min agoAmazon Bedrockzero-retention$1.00 in/1M$5.00 out/1M59 tok/s200K—30 min agoAmazon Bedrockzero-retention$1.10 in/1M$5.50 out/1M76 tok/s200K—30 min agoGoogle Vertex AIzero-retention$1.10 in/1M$5.50 out/1M76 tok/s200K—30 min agoGoogle Vertex AIzero-retention$1.10 in/1M$5.50 out/1M195 tok/s200K—30 min agoAmazon Bedrockzero-retention$1.10 in/1M$5.50 out/1M76 tok/s200K—30 min ago
Full specs, hardware verdicts and benchmarks on the Claude Haiku 4.5 model page →