LLM Pricing
ModelsLeaderboardsToolsProvidersReleasesArticles

© 2026 LLM Pricing

About
·Contact
·Privacy
·RSS
← Model list

Claude Opus 4.7

anthropic·anthropic/claude-opus-4-7·GA·Closed·claude-opus series
Stronger Opus tier for advanced software work and high-stakes reasoning

Specs & pricing

Input / output per 1M tokens
Official price·Anthropic
$5.00 / $25.00
Blended $10.00 · Cache read $0.50
Lowest paid·JZSTOKENGateway
$1.17 / $5.84
Blended $2.34 · 6.5× spread
1 more $0 channels
Context
1,000,000
Output limit
128,000
Knowledge cutoff
2026-01-31
Released / updated
2026-04-16 / 2026-04-16
Capabilities
✓ Reasoning✓ Tool use✓ Structured outputTemperature✓ Attachments
Modalities
TextImagePDF

Claude Opus 4.7 by anthropic is offered by 67 providers on this page. Public prices are shown for 65 of them. Its official list price is $5.00 per 1M input tokens and $25.00 per 1M output tokens. The lowest paid channel is JZSTOKEN at $1.17 / $5.84 per 1M, about 4.3× below the list price. That channel is a third-party gateway, so confirm its availability and rate limits before depending on it. It also has 1 free ($0) channel; free tiers usually carry rate limits, and subscription-covered access bills $0 per token only after the subscription fee.

Median output speed is 47 tokens per second, with 0.92s to the first token.

The context window is 1,000,000 tokens at the reference host, but hosts report different limits, from 200,000 to 1,048,576, so the usable window depends on the provider you pick. It supports reasoning, tool use, and structured output. Accepted input modalities are Text, Image, and PDF. Its training knowledge cuts off at 2026-01-31.

Quality & performanceArtificial Analysis · Intelligence Index v5 · rep. tier Non-reasoning, High

Intelligence—
Value—
Output speed47 tok/s
TTFT0.92 s

Quality is independently evaluated by Artificial Analysis. Speed/latency are model-level medians.

Available at 67 providers65 with public prices · 1 free

  • ProviderInput / Output
  • GitLab DuoCloud
    duo-chat-opus-4-7
    Free
    Cache read $0Context 1,000,000Output limit 128,000
  • JZSTOKENGateway
    $1.17 / $5.84
    Cache read $0.12Context 1,000,000Output limit 128,000
  • xxapiGateway
    $1.92 / $9.62
    Cache read $0.19Context 1,000,000Output limit 128,000
  • KoRouterGateway
    $3.25 / $16.25
    Cache read $0.33Context 1,000,000Output limit 128,000
  • KunavoGateway
    $3.50 / $17.50
    Cache read $0.35Context 1,000,000Output limit 128,000
  • PoeGateway
    anthropic/claude-opus-4.7
    $4.29 / $21.46
    Cache read $0.43Context 1,048,576 ⚠Output limit 128,000
  • GMI CloudCloud
    anthropic/claude-opus-4.7
    $4.50 / $22.50
    Cache read $0.45Context 409,600 ⚠Output limit 128,000
  • AnthropicOfficialFirst-party
    $5.00 / $25.00
    Cache read $0.50Context 1,000,000Output limit 128,000
  • Perplexity AgentGateway
    $5.00 / $25.00
    Cache read $0.50Context 1,000,000Output limit 128,000
  • CrossModelGateway
    $5.00 / $25.00
    Cache read $0.50Context 1,000,000Output limit 128,000
ProviderTierInputOutputCache readCache writeContextOutput limitStatus
GitLab Duo
duo-chat-opus-4-7
CloudFree$0$01,000,000128,000
JZSTOKENGateway$1.17$5.84$0.12$1.461,000,000128,000
xxapiGateway$1.92$9.62$0.19$2.401,000,000128,000
KoRouterGateway$3.25$16.25$0.33$4.061,000,000128,000
KunavoGateway$3.50$17.50$0.35$4.381,000,000128,000
Poe
anthropic/claude-opus-4.7
Gateway$4.29$21.46$0.43$5.371,048,576 ⚠128,000
GMI Cloud
anthropic/claude-opus-4.7
Cloud$4.50$22.50$0.45—409,600 ⚠128,000
AnthropicOfficialFirst-party$5.00$25.00$0.50$6.251,000,000128,000
Perplexity AgentGateway$5.00$25.00$0.50$6.251,000,000128,000
CrossModelGateway$5.00$25.00$0.50$6.251,000,000128,000

Sorted by blended price (input×0.75 + output×0.25) asc. The official channel always shows regardless of rank. Whether a gateway's low price is actually usable can't be verified.

Reasoning control

effort = loweffort = mediumeffort = higheffort = xhigheffort = maxeffort = noneToggle (on / off)budget_tokens ≥ 1,024

Interleaved thinking (reasoning between tool calls) is declared by 4 of 67 providers.

5 / 67 providers expose no reasoning control (reasoning_options: []).

Experimental modes6 items

ModeProviderInputOutputCache readCache write
fastGMI Cloud$30.00$150$3.00$37.50
fastMerge Gateway$30.00$150$3.00$37.50
fastGitHub Copilot$30.00$150$3.00$37.50
fastAuriko$30.00$150$3.00$37.50
fastOrcaRouter$30.00$150$3.00$37.50
fastAnyAPI$30.00$150$3.00$37.50

Your usage cost

Common scenarios
1JZSTOKEN$399.62
2xxapi$657.67
3KoRouter$1,111.50
4Kunavo$1,197.00
5Poe$1,468.18
6GMI Cloud$1,539.00
7Anthropic · Official$1,710.00

1 more channels offer $0 (GitLab Duo); free tiers usually have rate limits and no SLA, excluded from ranking.

Switch to JZSTOKEN to save $1,310.38/mo (77%)
Note: this is a gateway; verify availability and rate limits yourself.

Price historyone sample accumulated per data sync

Input listOutput listMin blended
$25.00$02026-08-052026-11-15
2026-11-15Input list $5.00Output list $25.00Min blended $2.34

Benchmark19 items

NameConditionsScoreMetricSource
SWE-Bench Pro—64.3resolve rateSource ↗
Terminal-Benchharness: Terminus-2 · v2.166.1success rateSource ↗
SWE-Atlas Refactoringharness: Claude Code48.57scoreSource ↗
Artificial Analysis Coding Agent Indexharness: Claude Code · variant: max66.6average pass@1Source ↗
SWE-Atlas Codebase QnAharness: Claude Code · variant: max81pass@1Source ↗
SWE-Bench Proharness: Claude Code · variant: max · dataset: hard-aa44.9pass@1Source ↗
Terminal-Benchharness: Claude Code · variant: max · v2.173.8pass@1Source ↗
Artificial Analysis Coding Agent Indexharness: Cursor CLI · variant: medium61.2average pass@1Source ↗
SWE-Atlas Codebase QnAharness: Cursor CLI · variant: medium78.4pass@1Source ↗
SWE-Bench Proharness: Cursor CLI · variant: medium · dataset: hard-aa34.4pass@1Source ↗
Terminal-Benchharness: Cursor CLI · variant: medium · v2.170.6pass@1Source ↗
Artificial Analysis Coding Agent Indexharness: Claude Code · variant: medium59.9average pass@1Source ↗
SWE-Atlas Codebase QnAharness: Claude Code · variant: medium71.7pass@1Source ↗
SWE-Bench Proharness: Claude Code · variant: medium · dataset: hard-aa36.4pass@1Source ↗
Terminal-Benchharness: Claude Code · variant: medium · v2.171.4pass@1Source ↗
GPQA Diamond—94.2accuracySource ↗
Humanity's Last Examvariant: no tools46.9accuracySource ↗
Humanity's Last Examvariant: with tools54.7accuracySource ↗
OSWorld-Verified—78success rateSource ↗

The same benchmark scores very differently across harness / dataset, so the qualifying conditions must be shown together.

Related models

Claude Opus 5.5same series$4.00 / $20.00Claude Opus Latestsame series—Claude Opus 5same series$5.00 / $25.00GLM-5.3cheaper alternative$1.40 / $4.40GLM-5.3-Flashcheaper alternative$0.15 / $0.50DeepSeek V4.1 Flash 1017cheaper alternative$0.15 / $0.60GLM-5.2cheaper alternative$1.40 / $4.40Claude Opus 4.6same lab$5.00 / $25.00Claude Opus 4.8same lab$5.00 / $25.00
Data partly from models.dev (MIT) · Anthropic official docs ↗