ModelRefs / Claude 3 Haiku - AI model implementation reference

Claude 3 Haiku - AI model implementation reference

Claude 3 Haiku is a retired Anthropic model — the fastest, most cost-efficient member of the original Claude 3 family, positioned for high-volume, latency-sensitive workloads with vision input support. Anthropic's model deprecations page lists claude-3-haiku-20240307 as deprecated on 19 February 2026 and retired on 20 April 2026, with Claude Haiku 4.5 as the recommended replacement. Those dates apply to Anthropic-operated platforms; Amazon Bedrock and Google Cloud set their own retirement schedules.

Overview

Claude 3 Haiku is attributed to Anthropic in ModelRefs' canonical registry. Tracked modalities: Text input and output, Image input. Primary use cases considered on ModelRefs: High-volume assistants, classification, and extraction with measured quality targets; Latency- and cost-sensitive vision-enabled workflows.

This ModelRefs profile is decision-support material, not a final or universal ranking. Confirm current behavior, access, pricing, limits, licensing, and lifecycle in Anthropic's own documentation, and evaluate Claude 3 Haiku on representative workloads before implementation.

Lifecycle

Retired. No longer served. Requests to it fail rather than falling back.

Deprecated
19 February 2026
Stopped serving
20 April 2026
Provider's replacement
claude-haiku-4-5-20251001
Applies to
Anthropic-operated platforms (Claude API, Claude Platform on AWS, Microsoft Foundry); Amazon Bedrock and Google Cloud set their own retirement schedules.

Provider-documented snapshots

  • claude-3-haiku-20240307

Source: platform.claude.com, checked 2026-10-16.

Benchmark & Evaluation

ModelRefs currently has partial, narrow benchmark coverage for Claude 3 Haiku. Treat the available benchmark evidence as one input to the decision, not a guarantee that Claude 3 Haiku is the strongest option for your workload, and evaluate it on representative workloads before selecting it.

  • No benchmark score is imported into this editorial record. Canonical benchmark runs and scores are governed separately with their own provenance and render only through those records; coverage in ModelRefs is currently narrow (partial), so any scored comparison must show its coverage limits.
  • ModelRefs holds canonical run evidence on graduate-level reasoning (GPQA); coverage is narrow and provider-reported evaluations are not independently reproduced here.

Implementation considerations

  • Benchmark Haiku directly against Sonnet-tier models on your task; the cheaper model is often sufficient but this must be measured.
  • Requests to claude-3-haiku-20240307 fail on Anthropic-operated platforms; Claude Haiku 4.5 should be benchmarked on its own rather than assumed equivalent.
  • Retired on Anthropic-operated platforms (Claude API, Claude Platform on AWS, Microsoft Foundry) since 20 April 2026; requests to a retired model ID fail rather than falling back.
  • Amazon Bedrock and Google Cloud publish their own model tables and retirement schedules, so availability there must be checked separately.
  • Feature flags, rate limits, and data-retention controls differ by platform and account; check the models overview for the current matrix.

Risks and limitations

  • Hosted-model behavior, quotas, pricing, and data controls can change without a client-side version pin unless a dated snapshot is used.
  • Provider-reported capabilities require task-specific evaluation before production reliance.

Source coverage

ModelRefs reviewed Anthropic's published material on 2026-07-12. Model behavior, access, pricing, limits, and lifecycle can change, so verify the linked provider documentation before implementation. What remains open is recorded on this page: Cost-quality routing guidance against higher Claude tiers needs task-level evidence. Fit and comparison signals remain provisional until tested on your own workload.

Connected ModelRefs evidence

Sources

Continue your research

Use these connected ModelRefs sections to compare alternatives, inspect implementation paths, and review the evidence and governance boundaries relevant to Claude 3 Haiku - AI model implementation reference.