Claude Opus 4.8
anthropic/claude-opus-4.8claude-opus-4-8claude-opus-4.8Claude-Opus-4.8opus-4-8opus-4.8- π’ Anthropic's most capable generally available Opus model, released May 2026.
- π 1M-token context window by default; 128K max output tokens.
- π§ Adaptive thinking, defaulting to high effort across all surfaces.
- π§ Better tool triggering and fewer agentic-coding compactions than Opus 4.7.
- ποΈ Multimodal: accepts text, image, and file inputs.
- π― Roughly four times less likely than predecessor to leave code flaws unremarked.
- β‘ Optional fast mode (research preview) reduces latency and cost.
- π Web search, function calling, and memory-driven long-horizon work.
Anthropic PBC is an American artificial intelligence company headquartered in San Francisco. Structured as a public benefit corporation, the lab develops large language models under the Claude name, with a research emphasis on building reliable, steerable, and safety-focused AIβ¦
Explore 18 more models by Anthropic βClaude Opus 4.8 is Anthropic's most capable generally available model, launched on May 28, 2026, and positioned for complex reasoning, long-horizon agentic coding, and high-autonomy knowledge work. It supports a 1M-token context window by default on the Claude API, Amazon Bedrock, and Vertex AI (200K on Microsoft Foundry), up to 128K output tokens, and adaptive thinking that defaults to high effort. It accepts text, image, and file inputs with text output.
The model is a direct upgrade to Claude Opus 4.7, itself following Claude Opus 4.6 and Claude Opus 4.5. Anthropic describes it as a "modest but tangible improvement," with gains concentrated in long-horizon agentic codingβbetter long-context handling, fewer compactions, and better compaction recoveryβplus improved tool triggering, addressing an issue some users reported on Opus 4.7. Anthropic reports the model is about four times less likely than Opus 4.7 to leave flaws in its own code unremarked, reflecting a focus on honesty and coherence over many turns.
On developer experience, the minimum cacheable prompt length drops to 1,024 tokens, and developers can now update instructions mid-task without breaking the prompt cache. These changes target the iterative, tool-heavy workflows the model is built for.
This release also introduces a fast-mode research preview, mirrored by the separate Claude Opus 4.8 Fast variant, alongside siblings including Claude Sonnet 4.6 and the newer Claude Fable 5.
This About section is AI-generated from public sources via VeniceStats + Venice inference, with no human editing. It may contain inaccuracies.
Daily usageΒ Β·Β Claude Opus 4.8Β Β·Β Organic providers
Top providersΒ Β·Β Claude Opus 4.8Β Β·Β Organic
- Venice.ai Proxy35.5%
- β² Apex Ant27.5%
- Super Seeder17.7%
- NovaRoute AI6.1%
- Open Bird5.0%
- antseed-neon-puma-944e1.6%
- Others6.8%
| Provider | Health | Repβ | Sales | Tokens | Routing | Input $/M | Cached $/M | Output $/M | Save | 1st token | tok/s |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Super Seeder 0xd19fβ¦41f3 | healthy | 90.57 | $54.50 | 14.73M | #3 | $4.7134 | $0.4713 | $23.567 | β6% | β | β |
| β² Apex Ant 0x73b4β¦e736 | healthy | 88.20 | $84.75 | 78.53M | #1 | $2.50 | $0.25 | $12.50 | β50% | β | β |
| Venice.ai Proxy 0x1f22β¦18c9 | healthy | 75.07 | $109.48 | 102.05M | #2 | $2.75 | $0.275 | $13.75 | β45% | β | β |
| Relay Works 0x3040β¦7139 | β | 58.03 | $0.25 | 178.28k | gated | $3.25 | $0.325 | $16.25 | β35% | β | β |
| ZLKPro-Api 0x0b0bβ¦f446 | healthy | 37.21 | $3.89 | 18.68M | gated | $0.094 | $0.05 | $0.284 | β98% | 1.6 s | 135 |
| Vikey Seller 0xe4c4β¦3c7b | healthy | 30.79 | $0.00 | 930.00 | gated | $2.50 | $0.25 | $12.50 | β50% | β | β |
| Prime Seed 0x71e2β¦fc07 | healthy | 27.20 | $1.36 | 243.98k | gated | $5.50 | $0.55 | $27.50 | β | β | β |
| HallPass 0xcac8β¦71b4 | β | 20.00 | β | β | gated | $0.55 | $0.08 | $2.65 | β89% | β | β |
| bartly.eth64.de 0x666eβ¦4666 | healthy | 7.22 | $2.80 | 3.43M | gated | $0.6865 | $0.0686 | $3.4324 | β86% | β | β |
| TokenSale 0x2060β¦742d | β | 5.30 | $0.22 | 1.71M | gated | $0.1259 | $0.039 | $0.6051 | β97% | β | β |
| RouteIT 0x64e5β¦b1df | healthy | 5.15 | $0.00 | 344.00 | gated | $0.65 | $0.075 | $3.00 | β87% | β | β |
| dealer 0xc46eβ¦8ce6 | β | 2.11 | $0.03 | 2.95k | gated | $4.75 | $0.475 | $23.75 | β5% | β | β |
| Hana Gateway β 0x4ae1β¦117b | healthy | 0.51 | $2.52 | 13.87M | gated | $0.14 | $0.02 | $0.65 | β97% | β | β |
| AI Inference Gateway 0xa40dβ¦8a15 | healthy | 0.00 | $0.00 | 213.00 | gated | $1,000.00 | $1,000.00 | $1,000.00 | β | β | β |
| antseed-seller-4 0xa2c6β¦d079 | β | 0.00 | β | β | gated | $10.00 | $0.00 | $20.00 | β | β | β |
| antseed-seller-5 0x4297β¦57ef | β | 0.00 | $0.71 | 957.01k | gated | $10.00 | $0.00 | $20.00 | β | β | β |
| antseed-seller-2 0x6df8β¦bfdb | β | 0.00 | β | β | gated | $10.00 | $0.00 | $20.00 | β | β | β |
| antseed-seller-5 0xa46eβ¦117f | β | 0.00 | β | β | gated | $10.00 | $0.00 | $20.00 | β | β | β |
| antseed-neon-raven-7393 0x28efβ¦bac9 | β | 0.00 | β | β | gated | $10.00 | $0.00 | $20.00 | β | β | β |
| antseed-seller-3 0x0d68β¦3fc3 | β | 0.00 | β | β | gated | $10.00 | $0.00 | $20.00 | β | β | β |
| antseed-seller-5 0x2de6β¦a09a | β | 0.00 | $0.00 | 408.00 | gated | $10.00 | $0.00 | $20.00 | β | β | β |
| antseed-seller-5 0x6b50β¦9fd1 | β | 0.00 | β | β | gated | $10.00 | $0.00 | $20.00 | β | β | β |
| antseed-seller-5 0x7357β¦abe5 | β | 0.00 | $0.03 | 2.59k | gated | $10.00 | $0.00 | $20.00 | β | β | β |
| antseed-seller-5 0xcc98β¦a38d | β | 0.00 | β | β | gated | $10.00 | $0.00 | $20.00 | β | β | β |
| surplus-hive 0x6145β¦5c58 | β | 0.00 | β | β | gated | $7.2917 | $7.2917 | $36.4583 | β | β | β |
| Open Ant 0xe4f6β¦5bc4 | β | 0.00 | $0.11 | 6.44k | gated | $6.00 | $0.60 | $30.00 | β | β | β |
| InferenceLab 0x74d4β¦2f72 | β | 0.00 | β | β | gated | $5.10 | $5.10not advertised | $25.10 | β | β | β |
| Fire Ant π₯π 0xbe05β¦bc5d | β | 0.00 | $0.02 | 10.00k | gated | $4.50 | $0.3563 | $22.50 | β10% | β | β |
| Apex TEE Test 0xe672β¦7955 | β | 0.00 | β | β | gated | $2.70 | $2.70 | $13.50 | β46% | β | β |
| LLMart 0x948aβ¦a000 | β | 0.00 | $0.17 | 131.84k | gated | $2.50 | $0.2557 | $12.50 | β50% | β | β |
| antseed-quiet-lynx-1012 0xef6eβ¦1012 | β | 0.00 | β | β | gated | $2.50 | $2.50not advertised | $12.50 | β50% | β | β |
| Leftermute 0x388bβ¦5389 | β | 0.00 | β | β | gated | $2.0786 | $2.0786not advertised | $10.3929 | β58% | β | β |
| antseed-neon-puma-944e 0x6650β¦944e | β | 0.00 | $4.79 | 593.26k | gated | $2.046 | $0.60 | $10.23 | β59% | β | β |
| Best 0x472fβ¦69fd | healthy | 0.00 | $2.54 | 16.82M | gated | $1.1629 | $0.9884 | $4.8371 | β77% | 0.3 s | 145 |
| Meridian AI 0x8c8cβ¦06f5 | β | 0.00 | $0.68 | 966.36k | gated | $1.0281 | $1.0281not advertised | $5.1403 | β79% | β | β |
| Xcat 0xbc33β¦b75f | β | 0.00 | $0.13 | 118.38k | gated | $1.00 | $1.00 | $5.00 | β80% | β | β |
| Open Bird 0xc0f1β¦8183 | β | 0.00 | $15.38 | 14.19M | gated | $0.7493 | $0.0749 | $3.7467 | β85% | β | β |
| Catastrophic 0xefbfβ¦8908 | β | 0.00 | β | β | gated | $0.35 | $0.25 | $0.95 | β93% | β | β |
| Flash 0x0329β¦9b2c | β | 0.00 | βexcluded | β | gated | $0.169 | $0.169not advertised | $0.99 | β97% | β | β |
| antseed-cosmic-gecko-c46d 0xd9c0β¦c46d | healthy | 0.00 | $0.27 | 1.91M | gated | $0.115 | $0.0805 | $0.345 | β98% | β | β |
| NovaRoute AI 0xc50dβ¦ed7b | β | 0.00 | $18.83 | 27.89M | gated | $0.1025 | $0.3663 | $0.3074 | β98% | β | β |
"Best price" and the provider table are live Antseed catalog data (advertised $/1M tokens β or $ per generated image for unit-billed image models β not settled amounts). Reputation = buyer trust score (0-100, the Antseed SDK's own formula). "Routing" = the SDK's default buyer routing (what Antseed Desktop ships with): a trust β₯ 60 gate on the effective reputation, then cheapest-first among routable providers; live failover state (per-peer cooldowns) is buyer-side runtime and not included. Model knowledge (TLDR, provider, About) via the VeniceStats enrichment layer. Advertised catalog, not the model used in any specific purchase. "Usage on Antseed" counts only settlements whose buyers share the per-model split on-chain (metadata v2/v3, opt-in), so every usage figure is a lower bound.