Claude Haiku 4.5
abacus/claude-haiku-4-5-20251001 · family claude-haiku · in: text, image → out: text · price provider-declared (via models.dev). Capabilities: tools reasoning text image.
$2.00
Blended $/1M (3:1)
$1
Input $/1M
$5
Output $/1M
$6.00
1M in + 1M out
—
Cache read $/1M
#5578 of 8,511
Cheapest rank
- Provider
- Abacus (
abacus) - Kind
- text (outputs text)
- Context
- 200,000 tokens
- Max output
- 64,000 tokens
- Capabilities
- tools reasoning text image
Same model, other providers
“Claude Haiku 4.5” is offered by 20 providers — cheapest-effective first. QiHang is the cheapest at $0.28 blended.
| Provider | Input | Output | Blended | Context |
|---|---|---|---|---|
| QiHang | $0.14 | $0.71 | $0.28 | 200k |
| Abacus (this) | $1 | $5 | $2.00 | 200k |
| Vertex | $1 | $5 | $2.00 | 200k |
| Vercel AI Gateway | $1 | $5 | $2.00 | 200k |
| Perplexity Agent | $1 | $5 | $2.00 | 200k |
| Tempr Gateway | $1 | $5 | $2.00 | 200k |
| CommandCode | $1 | $5 | $2.00 | 200k |
| FreeModel | $1 | $5 | $2.00 | 200k |
| Vertex (Anthropic) | $1 | $5 | $2.00 | 200k |
| Amazon Bedrock | $1 | $5 | $2.00 | 200k |
| FrogBot | $1 | $5 | $2.00 | 200k |
| Azure | $1 | $5 | $2.00 | 200k |
| ZenMux | $1 | $5 | $2.00 | 200k |
Call it
jishie indexes & prices models; it does not proxy inference.
Most providers are OpenAI-compatible — point base_url at Abacus and pass this model id:
curl $BASE_URL/v1/chat/completions \
-H "Authorization: Bearer $API_KEY" -H "Content-Type: application/json" \
-d '{"model":"claude-haiku-4-5-20251001","messages":[{"role":"user","content":"hello"}]}'
Route programmatically with MCP find_model / get_model, or fetch this record free (no wallet):