ANTHROPIC MODEL GUIDE
Claude Haiku 5.5 API
Claude Haiku 5.5 targets high-volume classification, data extraction, request routing, and small subtasks delegated by agents.
Core specificationsModel Plaza checked November 10, 2026
Model IDclaude-haiku-5-5
Context / max output1M / 128K
Input and outputText and image input; text output
ThinkingAdaptive thinking
Vendor statusActive
LLMFly AI input / output per 1M$0.05 / $0.25 (50% OFF)
Selection verdictClaude Haiku 5.5 bills at $0.05 input / $0.25 output per 1M tokens on LLMFly AI (0.5x standard tier).
Best fit
Test on short, repeatable requests and define how to flag ambiguous cases. Compare those cases with Claude Sonnet 5.5 before deciding when to escalate.
Avoid when
Complex, open-ended work without a way to detect and escalate uncertain results.
01
Claude API cost estimate
Enter token counts for one representative request; every usage item is multiplied by the request count.
Claude Code0.5xCurrent multiplier
Vendor reference cost$0.015Estimated savings$0.007550.0%
Cost breakdown
Rates are per 1M tokens- LLMFly AI
- $0.05
- Vendor reference
- $0.1
- Total tokens
- 100,000
- LLMFly AI
- $0.25
- Vendor reference
- $0.5
- Total tokens
- 10,000
- LLMFly AI
- $0.0625
- Vendor reference
- $0.125
- Total tokens
- 0
- LLMFly AI
- $0.005
- Vendor reference
- $0.01
- Total tokens
- 0
02
Choosing and integrating
Claude Haiku 5.5 API pricing breakdown
Claude Haiku 5.5 (claude-haiku-5-5) is billed by token usage. Prices below are USD per 1M tokens. LLMFly AI rates already include the 0.5x group multiplier; do not apply it again. Provider reference prices are separate. These rates come from the site's Model Plaza snapshot; confirm live catalog and account rates before calling.
Example: a request with 10,000 uncached input tokens and 1,000 output tokens costs approximately $0.00075 in LLMFly AI token charges. This excludes cache writes, tools, retries, and other additional fees.
- Platform and provider prices use their own input-token tiers. Do not apply provider long-context multipliers to platform prices.
- Not listed does not mean free. Account for cache reads, cache writes, and uncached input in their respective billing categories.
- Token cost = tokens in each billing category × its price per million ÷ 1,000,000. Add tools and other charges separately.
| Price source | Input tokens per request | Input | Output | Cache write | Cache read | 1h cache write |
|---|---|---|---|---|---|---|
| LLMFly AI | ≤100K | $0.05 | $0.25 | $0.0625 | $0.005 | $0.1 |
| LLMFly AI | >100K | $0.25 | $1.25 | $0.3125 | $0.025 | $0.5 |
| Provider reference | ≤100K | $0.1 | $0.5 | $0.125 | $0.01 | $0.2 |
| Provider reference | >100K | $0.5 | $2.5 | $0.625 | $0.05 | $1 |
Pricing tiers and additional charges
The lower tier includes requests with exactly 100,000 input tokens. Above that threshold, the higher rates apply to input, output, and cache usage for the full request.
Claude Haiku 5.5 capabilities and workloads
- The Claude API enables adaptive thinking by default. Recount tokens before migrating from Haiku 4.5 because the tokenizer changed.
Limits to check before migration
- Review old temperature, top_p, and top_k settings and omit them in native Claude requests. Check whether your existing prompts cross the new price threshold.
Comparing Claude Haiku 5.5 with sibling models
Test on short, repeatable requests and define how to flag ambiguous cases. Compare those cases with Claude Sonnet 5.5 before deciding when to escalate.
| Model ID | Context / max output | LLMFly AI in / out | Vendor reference in / out |
|---|---|---|---|
| claude-sonnet-5 | — | $1 / $5 | $2 / $10 |
| claude-sonnet-4-6 | 1M / 128K | $1.5 / $7.5 | $3 / $15 |
| claude-sonnet-4-5 | 200K / 64K | $1.5 / $7.5 | $3 / $15 |
| claude-haiku-5-5 | 1M / 128K | $0.05 / $0.25 | $0.1 / $0.5 |
| claude-haiku-4-5 | 200K / 64K | $0.5 / $2.5 | $1 / $5 |
Use the exact model ID when connecting
Copy the Claude Code setup from the console’s Use Key dialog, then run this command to test a text response.
- Send a minimal text request and inspect the response and usage.
- Test streaming, tool calls, and structured output separately if your application needs them.
- Check retry behavior and spending limits before increasing traffic.
claude -p "Reply with OK." --model claude-haiku-5-5Vendor capability versus gateway access
Claude Haiku 5.5 runs under the Claude Code group (0.5x multiplier) in Model Plaza. Configure Claude Code or Anthropic-compatible clients using the console's generated environment setup, and reconcile input, output, and cache line items in your usage log before scaling traffic.
Frequently asked questions
What should I put in the model field?
Use the exact ID claude-haiku-5-5, not the display name.
How is the discount calculated?
The current 0.5x multiplier is applied to the base rate, giving 50% off the catalog's official reference. Amounts are shown per million tokens.
Do these prices include every possible fee?
Not necessarily. Check cache-write duration, tools, and other route charges separately. Use the full pricing table and actual usage records to reconcile a bill.
What are the context and output limits?
Vendor documentation lists 1M context and 128K maximum output; confirm gateway route limits separately.
Use the Claude Haiku 5.5 API
LLMFly AI provides this model. Use the exact ID claude-haiku-5-5 and confirm current account pricing in Model Plaza.