anthropic/claude-haiku-4-5

The next generation of Anthropic's fastest and most cost-effective model, optimal for use cases where speed and affordability matter.

Use this model Compare providers

Modalities: TextImageText
In / Out / 1M: $1.20; $6.00 out
Context: 200K; tokens
Released: May 13, 2026

One model id, the best provider every call

Request anthropic/claude-haiku-4-5 and NeuralRouter scores every provider below on price, latency and health, then routes each request to the best fit — with automatic failover and unified billing.

Providers(1)

Provider	Input /1M	Output /1M	Cache read /1M	TTFT	Throughput	Uptime	Status
DeepInfradeepinfra/anthropic/claude-haiku-4-5	$1.20	$6.00	—cached	708 msp95	39 tok/s	100.0%30d	healthy	paymentsCheapestboltFastest

Effective pricing

Input

$1.20

Output

$6.00

Cache read

—

not offered

Best price shown (from DeepInfra). Each provider's full rate is in the table above — you always pay the provider the router picks.

Performance

Time to first tokenlower is better

DeepInfra

708 ms

Throughputhigher is better

DeepInfra

39 tok/s

Fastest provider responds in 708 ms to first token. Time-series charts build up as the probe runs on a schedule.

Benchmarks

No benchmark scores reported for this model by its providers.

Activity

Usage analytics appear here as requests flow through this model.

Call it from any OpenAI client

Point your base URL at NeuralRouter and request this model — we handle provider selection, failover and billing on every call.

# OpenAI-compatible — same SDK, smarter routing
curl https://api.neuralrouter.ai/v1/chat/completions \
  -H "Authorization: Bearer $NR_KEY" \
  -d '{"model":"anthropic/claude-haiku-4-5","messages":[{"role":"user","content":"Hello"}]}'