Claude Opus 4.8 · Anthropic

What does Claude Opus 4.8 cost to run in a coding agent?

Claude Opus 4.8 costs about $0.16 per average coding-agent request on our measured traffic, at list prices of $5 per million input and $25 per million output tokens. It scores 42 on the Artificial Analysis Intelligence Index, 16th of the 48 models we track. Source: Artificial Analysis (artificialanalysis.ai)

List prices: Anthropic pricing, checked 2026-10-13. Request costs: Standard Compute gateway data, 79,301 coding-agent requests from 75 users, Sep 13 to Oct 12, 2026. How we compute this

A newer version is out. Claude Opus 5.5 →

GitHub Copilot retires Claude Opus 4.8 on November 16, 2026, except for individual subscribers on annual plans. It stays on Anthropic's API. What changes and how to keep it →

Run Claude Opus 5.5 in your agent

Or let the router send each step to the cheapest model that can do it, on providers we vet for price, uptime and speed. Plans from $19 a month.

See plans

What Opus 4.8 costs per coding-agent request

Cost at list price	Opus 4.8
Prompt under 100K tokens50% of requests, 87% of input cached	$0.079
Prompt of 100K to 272K39% of requests, 88% of input cached	$0.20
Prompt over 272K tokens11% of requests, 86% of input cached	$0.43
Average request	$0.16
100 requests	$16.23
Median session16 requests; a session ends after 10 idle minutes	$2.25
A month of regular usemedian of 20 users active on 5+ days	$249

Each logged request priced at the vendor’s list rates for its own token counts, including cache reads and any higher rate for long prompts. Average output: 698 tokens per request.

Opus 4.8 benchmarks, speed and latency

Artificial Analysis	Value	Rank
Intelligence Index10 evaluations, 0 to 100	42	16 of 48
Terminal-Bench 4.0agentic coding in a terminal	21.7%	18 of 39
Output speedmedian	62 tokens/s	36 of 44
Time to first answer tokenincludes thinking time	66 s	24 of 44

Source: Artificial Analysis (artificialanalysis.ai). Measured variant: Claude Opus 4.8 (Max), speed on Anthropic. Fetched October 13, 2026. Rank is among the models on this site, older versions included. Artificial Analysis runs Terminal-Bench 4.0 in its own setup, so its scores differ from the tbench.ai leaderboard.

About Claude Opus 4.8

Claude Opus 4.8 was Anthropic's flagship from May 28, 2026 until Opus 5.5 replaced it in September. It has a 1M-token context and handles long tool-call chains, but it now trails the current Claude models clearly: 42 on the Artificial Analysis Intelligence Index and 21.7% in Artificial Analysis's Terminal-Bench 4.0 run, against 58 and 59.6% for Opus 5.5. It also costs more, at $5/$25 per million tokens against $4/$20, which works out to $0.16 per request against $0.107 in our measurements.

Sources: Artificial Analysis

Claude Opus 4.8 FAQ

How much does Claude Opus 4.8 cost per coding-agent request?

$0.16 on average at list price, $16.23 per 100 requests, and $2.25 for a median 16-request session. That is 79,301 real coding-agent requests (Sep 13 to Oct 12, 2026) priced at Anthropic's list rates; the median request sent 99,534 input tokens, 87% of input was read from cache, and requests returned 698 output tokens on average.

Does Claude Opus 4.8 charge more for long prompts?

No. Claude Opus 4.8 bills the same rate per token up to its 1M-token context window.

How fast is Claude Opus 4.8?

Artificial Analysis measures Claude Opus 4.8 (Max) at 62 output tokens per second, with 66 seconds to the first answer token, thinking time included, on Anthropic's API. Source: Artificial Analysis (artificialanalysis.ai).

Is Claude Opus 4.8 good for AI agents?

Existing Opus 4.8 setups until they move to Opus 5.5. On our scoring, which is calibrated to public benchmarks, it rates 6.7/10 for agentic ability and 8.0/10 for output quality. 1M-token context for whole-codebase work.

What is Claude Opus 4.8's context window?

Claude Opus 4.8 has a 1M-token context window, large enough to hold sizeable codebases or document sets in a single run. Anthropic is the maker.

Can I self-host Claude Opus 4.8?

No. Claude Opus 4.8 is a proprietary, hosted-only model from Anthropic. You access it through an API, such as OpenRouter, rather than running the weights yourself.

What are the downsides of Claude Opus 4.8?

The main trade-offs: Superseded by Claude Opus 5.5, which costs less ($4/$20 vs $5/$25) and scores far higher; Haiku 5.5 now scores higher on Artificial Analysis (43 vs 42) at a fraction of the price. It's strongest for existing Opus 4.8 setups until they move to Opus 5.5.