AI Model Watch
Models / Anthropic

Claude Opus 4.8

GA

Most capable Opus-tier model (complex reasoning, long-horizon agentic coding). API ID == alias 'claude-opus-4-8' (dateless pinned snapshot). Adaptive thinking; no extended thinking. effort defaults to 'high'. 5m cache $6.25/MTok, 1h cache $10/MTok, cache hits $0.50/MTok. Batch: $2.50 in / $12.50 out. 1M context at standard pricing; 200k on Microsoft Foundry. Supports up to 300k output via Batch API beta header. Fast mode (preview): $10 in / $50 out per MTok. Reliable knowledge cutoff Jan 2026; training data cutoff Jan 2026. Tentative retirement not sooner than May 28, 2027. Explicit release da

Claude Opus 4.8 by Anthropic costs $5 per 1M input tokens and $25 per 1M output tokens ($10/1M blended), with a 1M (1.000.000-token) context window. It is generally available (GA), with a knowledge cutoff of Jan 2026.

Last verified: 13 Aug 2026 · sourced from official provider documentation

Provider
Anthropic
Status
GA
Input price
$5 / 1M tokens
Output price
$25 / 1M tokens
Cached input
$0.5 / 1M tokens
Blended price
$10 / 1M tokens
Context window
1.000.000 tokens (1M)
Max output
128.000 tokens
Modality
text, image, vision
Open weights
No — API only
Knowledge cutoff
2026-01
Released
—
API string
claude-opus-4-8

Source: Anthropic official documentation ↗

See also

Compare Claude Opus 4.8 head-to-head

Claude Opus 4.8 vs GPT-5.6 Sol→

$10 vs $11.25 blended /M · cross-provider

Claude Opus 4.8 vs Gemini 3.6 Flash→

$10 vs $3 blended /M · cross-provider

Claude Opus 4.8 vs Grok 4.5→

$10 vs $3 blended /M · cross-provider

Claude Opus 4.8 vs Mistral Large 3→

$10 vs $0.75 blended /M · cross-provider

Claude Opus 4.8 vs Amazon Nova Pro→

$10 vs $1.40 blended /M · cross-provider

Claude Opus 4.8 vs Qwen-Max (Qwen2.5-Max)→

$10 vs $2.80 blended /M · cross-provider

FAQ

Claude Opus 4.8 — questions & answers

How much does Claude Opus 4.8 cost?

Claude Opus 4.8 is priced at $5 per 1M input tokens and $25 per 1M output tokens ($10/1M blended at a 3:1 input-to-output ratio), with cached input at $0.5 per 1M tokens.

What is the context window of Claude Opus 4.8?

Claude Opus 4.8 has a 1.000.000-token context window (1M), with up to 128.000 output tokens per request.

Is Claude Opus 4.8 deprecated?

No — Claude Opus 4.8 is generally available (GA) and not currently scheduled for deprecation or retirement in our tracker.

Does Claude Opus 4.8 support image or vision input?

Yes — Claude Opus 4.8 accepts image input. Listed modalities: text, image, vision.

Can I self-host Claude Opus 4.8?

No — Claude Opus 4.8 is proprietary. Its weights are not publicly released, so it is available only through Anthropic's API (or hosted partners), not for self-hosting.

What is the API model string for Claude Opus 4.8?

The API model identifier for Claude Opus 4.8 is "claude-opus-4-8" when calling Anthropic's API.

What is Claude Opus 4.8's knowledge cutoff?

Claude Opus 4.8's training knowledge cutoff is Jan 2026.