Claude Fable 5 Launches with 1M Token Context
Anthropic released Claude Fable 5 on June 9, 2026. The model features a 1M token context window, 128k max output tokens, and always-on adaptive thinking for…
Official Release Details
Anthropic officially launched Claude Fable 5 on June 9, 2026. The announcement stated that this model is the provider's most capable widely released model. It was released alongside Claude Mythos 5, which is designated specifically for Project Glasswing participants. The source material confirms that both models share specific technical specifications regarding their processing limits.
- The model identifier is claude-fable-5.
- It supports a default context window of 1,000,000 tokens.
- The maximum output length is set at 128,000 tokens.
- Always-on adaptive thinking is included in the feature set.
Model Specifications and Access
According to the official changelog, Claude Fable 5 accepts text and image inputs while producing text outputs. The documentation notes that open weights are not confirmed for this release. Pricing details were referenced in the announcement but are not specified in the provided source facts. The model appears in the Anthropic models API, confirming its availability through standard API channels. No information regarding local installation requirements or specific deployment architectures is present in the verified data.
Evaluation Procedure for Engineers
Engineers can evaluate the model by testing its performance against specific use cases. To validate the context window capability, users should input documents totaling close to one million tokens and measure the model's ability to retrieve relevant information from distant parts of the prompt. To compare output quality, developers should generate responses using the 128k max output limit and assess the coherence of long-form text generation. These steps allow for a practical assessment of the documented features without assuming unverified benefits.
Artificial Analysis benchmark review
Benchmark review updated on . These scores apply to the linked artifact and recorded methodology; they do not establish the performance of a different runtime or quantized model.
| Measurement | Recorded result | Measured |
|---|---|---|
| Artificial Analysis Coding Index (v4.1) ↗ | 76.5 index_points | 2026-08-09 |
| Artificial Analysis Coding Index (v4.2) ↗ | 76.5 index_points | 2026-09-05 |
| Artificial Analysis Coding Index (v4.3) ↗ | 76.5 index_points | 2026-09-20 |
| Artificial Analysis Intelligence Index (v4.1) ↗ | 62.1 index_points | 2026-08-09 |
| Artificial Analysis Intelligence Index (v4.2) ↗ | 53.2 index_points | 2026-09-05 |
| Artificial Analysis Intelligence Index (v4.3) ↗ | 49.6 index_points | 2026-09-20 |
Review the full benchmark provenance and source methodology →
Primary sources
Frequently asked questions
What is the exact date of the Claude Fable 5 release?
The source facts state that the release occurred on June 9, 2026. This is the historical date of the event, distinct from the current publication date of this article.
Does the source confirm if Claude Fable 5 has open weights?
No. The source records that open weights are not confirmed. It does not establish whether local installation is possible or impossible, nor does it specify any hosting arrangements beyond API observation.