How much does Claude Opus 4.8 cost?

On the Respan gateway, Claude Opus 4.8 costs from $2.50 per 1M input tokens and $12.50 per 1M output tokens on OpenRouter, after gateway discounts. Averaged across its 10 routes from 8 providers, it costs $4.80 per 1M input tokens and $24.00 per 1M output tokens.

Which providers serve Claude Opus 4.8 on Respan?

Claude Opus 4.8 is available through 10 routes from 8 providers: OpenRouter, Anthropic, Bedrock, DeepInfra, Inference.net, Perplexity, and 2 more. Bedrock, Tera, and Google Vertex AI require your own provider API key (BYOK).

What is Claude Opus 4.8's context window?

Claude Opus 4.8 accepts up to 1M tokens of context on its largest provider route.

How do I use Claude Opus 4.8 with the Respan gateway?

Point any OpenAI-compatible SDK at https://api.respan.ai/api/ with your Respan API key and set the model to "anthropic/claude-opus-4-8". To set it up with a coding agent, copy the agent setup prompt at the top of this page.

How fast is Claude Opus 4.8?

Over the last 30 days of requests through the Respan gateway, the fastest provider was Anthropic at 65 tokens per second, and the lowest response latency was 0.8 seconds on Bedrock. Speed varies with request size and provider.

Where does Claude Opus 4.8 rank on the Respan Index?

Claude Opus 4.8 scores 52.1 out of 100 on the Respan Index v1.1, ranking #20 of 143 models. By category: Coding 35.3, Agents 64.3, Reasoning 55.2, Knowledge & language 52.5, and Human preference 54.3. The Respan Index combines independent and lab-reported benchmark results into one score, using benchmark data collected by llmmetric.com.

How does Claude Opus 4.8 score on benchmarks?

Claude Opus 4.8 has 41 published benchmark scores, including SWE-bench Verified 88.6%, BrowseComp 84.3%, and GPQA Diamond 93.6%. Each score links to where it was published; the data is collected by llmmetric.com.

Claude Opus 4.8

Q: What is Claude Opus 4.8?

Claude Opus 4.8 is a model from Anthropic, released on May 28, 2026. Claude Opus 4.8 is Anthropic's most capable generally available model in the Opus family. It supports text, image, and file inputs with text output, with reasoning support and a 1M-token context window. It is suited for highly autonomous agents, long-horizon agentic work, knowledge work, and memory-driven tasks where coherence over extended sessions matters. It is particularly strong on multi-step reasoning, complex coding, and end-to-end project orchestration - large codebases, multi-stage debugging, and long-running asynchronous agent pipelines. Beyond coding, it handles knowledge work such as drafting documents, building presentations, and analyzing data, maintaining quality across very long outputs.

Claude Opus 4.8 is Anthropic's most capable generally available model in the Opus family. It supports text, image, and file inputs with text output, with reasoning support and a 1M-token context window. It is suited for highly autonomous agents, long-horizon agentic work, knowledge work, and memory-driven tasks where coherence over extended sessions matters. It is particularly strong on multi-step reasoning, complex coding, and end-to-end project orchestration - large codebases, multi-stage debugging, and long-running asynchronous agent pipelines. Beyond coding, it handles knowledge work such as drafting documents, building presentations, and analyzing data, maintaining quality across very long outputs.

Input / output, from: $2.50 / $12.50 per 1M
Weekly tokens / requests: 1.3K / 231
Context: 1M tokens
Released: May 28, 2026

Try Respan for free

Agent setup

Paste into Claude Code, Cursor, or Codex

Help me set up the Respan gateway: fetch and follow https://www.respan.ai/docs/agent-setup.md. Use model "anthropic/claude-opus-4-8".

from openai import OpenAI

client = OpenAI(

    base_url="https://api.respan.ai/api/",

    api_key="YOUR_RESPAN_API_KEY",

)

response = client.chat.completions.create(

    model="anthropic/claude-opus-4-8",

    messages=[{"role": "user", "content": "Hello!"}],

)

print(response.choices[0].message.content)

import OpenAI from "openai";

const client = new OpenAI({

  baseURL: "https://api.respan.ai/api/",

  apiKey: "YOUR_RESPAN_API_KEY",

});

const response = await client.chat.completions.create({

  model: "anthropic/claude-opus-4-8",

  messages: [{ role: "user", content: "Hello!" }],

});

console.log(response.choices[0].message.content);

curl https://api.respan.ai/api/chat/completions \

  -H "Authorization: Bearer YOUR_RESPAN_API_KEY" \

  -H "Content-Type: application/json" \

  -d '{

    "model": "anthropic/claude-opus-4-8",

    "messages": [{"role": "user", "content": "Hello!"}]

}'

Read docs

Providers

10 routes from 8 providers, cheapest first, priced per 1M tokens after gateway discounts. BYOK routes use your own key.

Provider	Context	Input	Output	Cache read	Capabilities
OpenRouter`openrouter/anthropic/claude-opus-4.8:batch`	1M	$2.50/M	$12.50/M	$0.25/M	+2
Anthropic`anthropic/claude-opus-4-8`	1M	$5.00/M	$25.00/M	$0.50/M	+2
BedrockBYOK only`bedrock/global.anthropic.claude-opus-4-8`	1M	$5.00/M	$25.00/M	$0.50/M	+2
DeepInfra`deepinfra/anthropic/claude-opus-4-8`	1M	$5.00/M	$25.00/M	$0.00/M	+2
Inference.net`inference_net/claude-opus-4-8`	1M	$5.00/M	$25.00/M	$0.50/M
OpenRouter`openrouter/anthropic/claude-opus-4.8`	1M	$5.00/M	$25.00/M	$0.50/M	+2
Perplexity`perplexity/anthropic/claude-opus-4-8`	1M	$5.00/M	$25.00/M	$0.50/M
TeraBYOK only`tera/anthropic/claude-opus-4-8`	200K	$5.00/M	$25.00/M	$0.50/M	+2
Google Vertex AIBYOK only`vertex_ai/claude-opus-4-8@default`	200K	$5.00/M	$25.00/M	$0.50/M	+2
BedrockBYOK only`bedrock/us.anthropic.claude-opus-4-8`	1M	$5.50/M	$27.50/M	$0.55/M	+2

Is a provider down? OpenRouter status, Anthropic status, AWS Bedrock status, Perplexity status, Google AI status

Pricing

Averaged across 10 routes from 8 providers, in USD per 1M tokens after gateway discounts.

Average input: $4.80 per 1M
Average output: $24.00 per 1M
Average cache read: $0.43 per 1M
Average cache write: $4.13 per 1M

Cost by volume

Total cost from 0 to 1T tokens on the cheapest and most expensive provider.

Highest$5.50/MBedrockLowest$2.50/MOpenRouter

Throughput and latency

30-day averages from requests through the Respan gateway, 3 of 10 routes had traffic in that window. Under each figure, every model on the gateway by its average, this one in blue.

Output speed

Higher is better

65tokens/sbest, on Anthropic

Faster than 71% of the other 485 models on the gateway

Anthropic65 tok/s
Perplexity8.8 tok/s
Bedrock5.3 tok/s

Time to first token

Lower is better

1.1sbest, on Perplexity

Faster than 64% of the other 64 models on the gateway

Perplexity1.1 s
Anthropic1.6 s

Response latency

Lower is better

0.8sbest, on Bedrock

Faster than 13% of the other 501 models on the gateway

Bedrock0.8 s
Perplexity1.5 s
Anthropic10 s

No traffic in the last 30 days on OpenRouter, Bedrock, DeepInfra, Inference.net, Tera, and Google Vertex AI.

Benchmarks

The Respan Index plus 41 published scores from 14 sources, each linked to where it was published.

52.1/ 100

Respan Index#20 of 143

143 ranked models. The shaded band is this model's 95% interval, 47.6 to 56.5. GPT-5.5 and Grok 4.6 fall inside it, so the gap is within the margin of error.

20Claude Opus 4.852.1
19GPT-5.552.5
21Grok 4.651.1

Published scores

Benchmark	Score	Source	Published
SWE-bench VerifiedAnthropic	88.6%	Anthropic	May 28, 2026
SWE-bench MultilingualAnthropic	84.4%	Anthropic	May 28, 2026
WeirdMLHåvard Tveit Ihle	82.9%	Håvard Tveit Ihle	Oct 5, 2026
LiveBench CodingLiveBench	81.8%	LiveBench	Oct 5, 2026
Terminal-Bench 2.1Anthropic	74.6%	Anthropic	May 28, 2026
SWE-Bench ProAnthropic	69.2%	Anthropic	May 28, 2026
DeepSWE v1.1Datacurve	59%	Datacurve	Oct 5, 2026
GSO Opt@1GSO	47.1%	GSO	Jul 12, 2026
SWE-bench MultimodalAnthropic	38.4%	Anthropic	May 28, 2026
Terminal-Bench 4.0Vals AI	23.2%	Vals AI	Oct 5, 2026

Benchmark	Score	Source	Published
BrowseCompAnthropic	84.3%	Anthropic	May 28, 2026
OSWorld-VerifiedAnthropic	83.4%	Anthropic	May 28, 2026
Toolathlon-VerifiedHKUST	76.2%	HKUST	Jun 30, 2026
LiveBench Agentic CodingLiveBench	50.5%	LiveBench	Oct 5, 2026
APEX-AgentsMercor	48.9%	Mercor	Oct 5, 2026
tau2-bench Banking KnowledgeSierra	39.7%	Sierra	Aug 4, 2026
Vending-Bench 2Andon Labs	5787.43 · #21 of 59	Andon Labs	Oct 5, 2026

Benchmark	Score	Source	Published
GPQA DiamondAnthropic	93.6%	Anthropic	May 28, 2026
GPQA DiamondEpoch AI	91%	Epoch AI	Oct 5, 2026
LiveBench ReasoningLiveBench	89.2%	LiveBench	Oct 5, 2026
LiveBench Data AnalysisLiveBench	66%	LiveBench	Oct 5, 2026
SimpleBenchSimpleBench	64.8%	SimpleBench	May 29, 2026
Humanity's Last Exam (with tools)Anthropic	57.9%	Anthropic	May 28, 2026
Humanity's Last Exam (no tools)Anthropic	49.8%	Anthropic	May 28, 2026
Mystery Game PuzzlesEpoch AI	36%	Epoch AI	Oct 5, 2026
Chess PuzzlesEpoch AI	34%	Epoch AI	Oct 5, 2026
EBR-benchEpoch AI	28.6%	Epoch AI	Oct 5, 2026

Benchmark	Score	Source	Published
AIME 2026MathArena	100%	MathArena	Sep 17, 2026
OTIS Mock AIME 2024-2025Epoch AI	98.3%	Epoch AI	Oct 5, 2026
LiveBench MathematicsLiveBench	94.3%	LiveBench	Oct 5, 2026
FrontierMath Tiers 1-3 v2Epoch AI	80%	Epoch AI	Oct 5, 2026
FrontierMath Tier 4 v2Epoch AI	56.1%	Epoch AI	Oct 5, 2026

Benchmark	Score	Source	Published
LiveBench LanguageLiveBench	79.7%	LiveBench	Oct 5, 2026
LiveBench Instruction FollowingLiveBench	72%	LiveBench	Oct 5, 2026
SimpleQA VerifiedEpoch AI	53%	Epoch AI	Oct 5, 2026

Benchmark	Score	Source	Published
Furniture AssemblyEpoch AI	42.5%	Epoch AI	Oct 5, 2026

Benchmark	Score	Source	Published
LMArena WebDevLMArena	1556 Elo · #32 of 99	LMArena	Oct 8, 2026
LMArena EloLMArena	1481 Elo · #22 of 147	LMArena	Oct 8, 2026
LMArena VisionLMArena	1286 Elo · #18 of 85	LMArena	Oct 9, 2026
LMArena AgentLMArena	0.0776 · #10 of 51	LMArena	Oct 12, 2026