Cost per coding-agent request

Is Claude Opus 4.8 or Devstral 2 cheaper for coding agents?

For coding agents, Claude Opus 4.8 costs 13x as much as Devstral 2 per request ($0.13 vs $0.010 on our measured traffic), because every Claude Opus 4.8 rate is 13 times Devstral 2's.

List prices: Anthropic pricing, OpenRouter (vendor endpoint), 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

Run Claude Opus 5.5 in your agent from $19/month

See plans

Cost per coding-agent request, side by side

Cost at list price	Devstral 2	Opus 4.8
Prompt under 100K tokens50% of requests, 87% of input cached	$0.0062	$0.079
Prompt of 100K to 272K39% of requests, 88% of input cached	$0.015	$0.19
Prompt over 272K tokens11% of requests, 86% of input cached	n/a	n/a
Average request	$0.010	$0.13
100 requests	$1.01	$12.72
Median session16 requests; a session ends after 10 idle minutes	$0.12	$1.54
A month of regular usemedian of 20 users active on 5+ days	$12.73	$161

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. Both models are priced on the requests up to 262K tokens that fit in both context windows; band shares above describe all requests.

Benchmarks, speed and latency, side by side

Artificial Analysis	Devstral 2	Opus 4.8
Intelligence Index10 evaluations, 0 to 100	9	42
Terminal-Bench 4.0agentic coding in a terminal	0.0%	21.7%
Output speedmedian	n/a	62 tokens/s
Time to first answer tokenincludes thinking time	n/a	66 s

Source: Artificial Analysis (artificialanalysis.ai). Measured variants: Devstral 2; Claude Opus 4.8 (Max), speed on Anthropic. Fetched October 13, 2026. Artificial Analysis runs Terminal-Bench 4.0 in its own setup, so its scores differ from the tbench.ai leaderboard.

What to watch for

Devstral 2's 262K window cannot take 12% of coding-agent requests, so the comparison uses only requests both can take. Claude Opus 4.8 is superseded: Claude Opus 5.5 costs $0.11 per request. Claude Opus 4.8 scores clearly higher on the benchmarks above, so it may finish hard tasks in fewer steps.

FAQ

Is Devstral 2 cheaper than Claude Opus 4.8 for coding agents?

Yes, at list prices. 100 average coding-agent requests cost $1.01 on Devstral 2 and $12.72 on Claude Opus 4.8. A median 16-request session costs $0.12 vs $1.54. Measured on 79,301 real coding-agent requests (Sep 13 to Oct 12, 2026), counting only the requests up to 262K tokens that both can take.

How much does a month of Opus 4.8 vs Devstral 2 cost?

For the median regular user in our data (1,478 requests in 30 days), about $161 on Claude Opus 4.8 and $12.73 on Devstral 2 at list price. Heavier users pay proportionally more.

Claude Opus 4.8 vs Devstral 2: which is better for agents?

Claude Opus 4.8 scores higher on output quality (8.0/10 vs 3.3/10), while Devstral 2 is the cheaper option. Claude Opus 4.8 is best for existing Opus 4.8 setups until they move to Opus 5.5; Devstral 2 is best for coding agents that need a Mistral open-weight model. Live community votes on this page show what people actually prefer.

Can I switch between Claude Opus 4.8 and Devstral 2 in the same agent?

Yes. Model-agnostic agents (Cline, Aider, OpenClaw, Hermes Agent, Kilo Code, OpenCode, Continue) let you swap models per task. Many people route hard tasks to the stronger model and routine work to the cheaper one.