ReasoningStandard Reason 72Bstandard-reason-72b
128KCONTEXT
112 msMEDIAN
94 tok/sPEAK
$0.90 / $3.60IN / OUT
Live↗
The standard for
all intelligence.
ONE COMPATIBLE ENDPOINTA model platform for every form of intelligence—from fast text and deep reasoning to vision and retrieval.
AVAILABLE MODELS24ONLINE
FASTEST TTFT36MS
MAX CONTEXT128K TOKENS
API COMPATIBILITYOpenAIFORMAT
Find the right
model.
Every model below is available through the same base URL. Switch model IDs without rebuilding your integration.
Prices are shown per one million tokens. Latency is measured on the shared production cluster.
ReasoningStandard Reason 32Bstandard-reason-32b
128KCONTEXT
86 msMEDIAN
128 tok/sPEAK
$0.55 / $2.20IN / OUT
Live
CodeStandard Code 34Bstandard-code-34b
64KCONTEXT
74 msMEDIAN
146 tok/sPEAK
$0.40 / $1.60IN / OUT
Live
Fast textStandard Flash 8Bstandard-flash-8b
64KCONTEXT
48 msMEDIAN
218 tok/sPEAK
$0.15 / $0.60IN / OUT
Live
MultimodalMIRA Vision 32Bmira-vision-32b
128KCONTEXT
146 msMEDIAN
76 tok/sPEAK
$0.50 / $1.80IN / OUT
Live
MultimodalMIRA Vision 11Bmira-vision-11b
64KCONTEXT
92 msMEDIAN
119 tok/sPEAK
$0.28 / $1.05IN / OUT
Live
EmbeddingECHO Embed 3echo-embed-3
8KCONTEXT
22 msP50
4,096DIMENSIONS
$0.08INPUT
Live
RerankerECHO Rerank 1echo-rerank-1
32KCONTEXT
31 msP50
1K docsPER REQUEST
$0.12INPUT
Live
Need a private or fine-tuned model?
Deploy isolated capacity with custom weights, security boundaries, and performance targets.