VENICE ATLAS · MODEL DOSSIER · CLAUDE OPUS 4.8

Model overview

Claude Opus 4.8

Name: Claude Opus 4.8
Brand: venice.ai
Price: 5.0000 USD
Availability: InStock

Claude Opus 4.8 is Anthropic's most capable generally available model in the Opus family. It supports long-horizon agentic work, complex multi-step coding, and memory-driven tasks where coherence over extended sessions matters. It features a 1M token context window, 128K max output tokens, adaptive thinking, and strong multimodal capabilities.

claude-opus-4-8 · venice.ai

Text · Venice catalog$5.00 per 1M tokensPrice on Venice AI

ModalityTextOn Venice AI

Primary price$5.00 per 1M tokensper 1M tokens

Context window1MMax tokens

Benchmarks27Benchmark results

Benchmark results

See the tests behind each score. They use the benchmark’s original scoring scale.

Agents#2433.79 rating Biology#638.39 rating Chemistry#838.39 rating Coding#1338.70 rating Vision#836.83 rating Frontend Development#232.61 rating General intelligence#1749.75 rating Grounding#224.63 rating Long Context#1720.42 rating Math#1239.59 rating Multimodal#836.14 rating Physics#738.39 rating Reasoning#1349.82 rating Safety#1410.22 rating Search#727.73 rating Spatial Reasoning#728.47 rating Tool Calling#1627.89 rating

Agents4 results

Finance Agentfinance-agent · agents

53.9%

ReportedOct 2026

Finance Agent v2finance-agent-v2 · agents

53.9%

ReportedOct 2026

OfficeQA Proofficeqa-pro · agents

66.2%

ReportedOct 2026

Toolathlontoolathlon · agents

59.9%

ReportedOct 2026

Coding3 results

CyberGymcybergym · code

78.8%

ReportedOct 2026

DeepSWE 1.1deepswe-1.1 · code

59.0%

ReportedOct 2026

FrontierSWEfrontierswe · code

75.0%

ReportedOct 2026

General intelligence1 result

Includeinclude · general

87.6%

ReportedOct 2026

Healthcare1 result

HealthBench Professionalhealthbench-professional · healthcare

55.8%

ReportedOct 2026

Math1 result

LiveBenchlivebench · math

77.2%

ReportedOct 2026

Multimodal1 result

OSWorld-Verifiedosworld-verified · multimodal

83.4%

ReportedOct 2026

Reasoning14 results

BrowseCompbrowsecomp · reasoning

84.3%

ReportedOct 2026

CharXiv-Rcharxiv-r · reasoning

89.9%

ReportedOct 2026

FrontierCode 1.1frontiercode-1.1 · reasoning

46.5%

ReportedOct 2026

GPQAgpqa · reasoning

93.6%

ReportedOct 2026

Graphwalks BFS >128kgraphwalks-bfs->128k · reasoning

68.1%

ReportedOct 2026

Graphwalks parents >128kgraphwalks-parents->128k · reasoning

83.3%

ReportedOct 2026

Humanity's Last Examhumanity's-last-exam · reasoning

57.9%

ReportedOct 2026

MCP Atlasmcp-atlas · reasoning

82.2%

ReportedOct 2026

SWE-bench Multilingualswe-bench-multilingual · reasoning

84.4%

ReportedOct 2026

SWE-Bench Multimodalswe-bench-multimodal · reasoning

38.4%

ReportedOct 2026

SWE-Bench Proswe-bench-pro · reasoning

69.2%

ReportedOct 2026

SWE-Bench Verifiedswe-bench-verified · reasoning

88.6%

ReportedOct 2026

Terminal-Bench 2.0terminal-bench-2 · reasoning

74.6%

ReportedOct 2026

Terminal-Bench 4.0terminal-bench-4.0 · reasoning

23.6%

ReportedOct 2026

Search1 result

DeepSearchQAdeepsearchqa · search

93.1%

ReportedOct 2026

Vision1 result

ScreenSpot Proscreenspot-pro · vision

87.9%

ReportedOct 2026

Learn what each benchmark measures ↗

Pricing detail

Input price per 1M tokens: $5.00
Output price per 1M tokens: $25.00
Cache read price per 1M tokens: $0.50
Cache write price per 1M tokens: $6.25
Blended price per 1M tokens (50/50): $15.00

Prices are published by Venice AI. Venice model page ↗

Capabilities

ReasoningFunction callingVisionWeb search

Endpoint metadata

Privacy: anonymized
Quantization: not-available
Maximum completion tokens: 128K
Uncensored: No
Beta: No
Mapping confidence: 100.0%

All capabilities

Max Images: 10
Quantization: not-available
Supports E2 E E: No
Supports Vision: Yes
Supports X Search: No
Optimized For Code: Yes
Supports Log Probs: No
Supports Reasoning: Yes
Supports Web Search: Yes
Supports Audio Input: No
Supports Video Input: No
Default Reasoning Effort: medium
Reasoning Effort Options: low, medium, high, xhigh, max
Supports Multiple Images: Yes
Supports Response Schema: Yes
Supports Tee Attestation: No
Supports Function Calling: Yes
Supports Reasoning Effort: Yes

Related models

More from anthropic

claude-sonnet-5-5Claude Sonnet 5.5$2.00 input per 1M tokens ↗claude-haiku-5-5Claude Haiku 5.5$0.10 input per 1M tokens ↗claude-opus-5-5Claude Opus 5.5$4.00 input per 1M tokens ↗