Fugu Ultra vs Claude Fable 5
Benchmarks, API Token Pricing, Context Window & Speed Comparison
Related topics (12)
Sakana AI 的 Fugu Ultra 与 Anthropic 的 Claude Fable 5 是当前备受关注的两款主力 AI 模型。Fugu Ultra 提供 1M Token 上下文窗口，API 定价为每百万 Token 输入 $1.50 / 输出 $6.00；而 Claude Fable 5 提供 1M Token 上下文窗口，定价为每百万 Token 输入 $10.00 / 输出 $50.00。以下为两者的基准测试、延迟与成本全面对比。
切换并对比任意模型
Select any AI models from our verified benchmark database to generate live side-by-side matrices.
Fugu Ultra
深度逻辑推理引擎Sakana AI 推出的 Fugu Ultra 是一款专为博士级数理逻辑、科学推导与多步思维链 (CoT) 设计的深度推理大模型，具备 1M 上下文窗口、75 Token/秒生成速度与 86/100 综合智能指数（每百万 Token $1.5 输入 / $6 输出）。
$1.50 / 1M
$6.00 / 1M
1M tokens
N/A
Claude Fable 5
旗舰前沿大模型Anthropic 推出的 Claude Fable 5 是一款面向复杂企业级推理、软件开发与全场景生产工作流的旗舰前沿大模型，具备 1M 上下文窗口、52 Token/秒生成速度与 72/100 综合智能指数（每百万 Token $10 输入 / $50 输出）。
$10.00 / 1M
$50.00 / 1M
1M tokens
76.5%
核心差异一览
- Fugu Ultra is 7.8x cheaper for API usage (input + output tokens combined).
- Fugu Ultra is faster at 75 tok/s vs 52 tok/s.
| Feature | Fugu Ultra | Claude Fable 5 |
|---|---|---|
Intelligence Index Composite reasoning & knowledge score (0-100) | 86 | 72 |
Provider | Sakana AI | Anthropic |
Input Price Cost per million input tokens | $1.50 / 1M | $10.00 / 1M |
Output Price Cost per million output tokens | $6.00 / 1M | $50.00 / 1M |
Blended Price Weighted 7:2:1 cache:input:output per 1M | $1.01 / 1M | $7.70 / 1M |
Context Window Maximum input + output tokens | 1M tokens~1,500 A4 pages of text | 1M tokens~1,500 A4 pages of text |
Max Output Tokens Per single request | 16K | 128K |
Throughput Speed Tokens generated per second | 75 tok/s | 52 tok/s |
SWE-bench Verified Real-world software engineering | N/A | 76.5% |
GPQA Diamond Graduate-level reasoning | 67% | 79.2% |
Cached Input Price Prompt caching discount per 1M tokens | $0.15 / 1M | $1.00 / 1M |
各项指标谁更胜一筹？
At-a-glance decision frameworkEvaluations ongoing
权威基准测试得分对比
Official evaluation metrics verified by independent third-party research labs.
Industry benchmark evaluating real-world GitHub issue resolution and multi-file debugging.
Challenging questions created by domain experts to test complex PhD-level reasoning.
Hardened reasoning and academic knowledge benchmark with reduced lucky-guess probability.
Crowdsourced human preference evaluation based on blind, head-to-head model battles.
Autonomous multi-step tasks executed inside a real command-line terminal environment.
Tokens emitted per second during active generation streaming.
延迟与响应速度
Time to first token (TTFT) and estimated end-to-end response time for a typical 500-token response.
实时 Token 成本模拟器
Adjust sliders to calculate projected monthly cloud API spending with prompt caching discounts.
Fugu Ultra · Strengths & Key Advantages
- 提供 1M 上下文窗口与最高 16K 输出容量，核心基准表现优异（GPQA Diamond 67% · MMLU-Pro 83%）
- 高达 75 Token/秒的生成吞吐量，首字延迟 (TTFT) 仅约 540 毫秒
- 极具竞争力的 API 定价（每百万 Token $1.5 输入 / $6 输出），原生支持提示词缓存、函数调用与结构化 JSON 输出
Considerations & Trade-offs
- 闭源专有云端模型，不支持下载权重进行离线本地物理隔离部署
- 在超长链路的前沿科研推导任务上，深度略逊于顶配重型旗舰推理模型
Claude Fable 5 · Strengths & Key Advantages
- 提供 1M 上下文窗口与最高 128K 输出容量，核心基准表现优异（SWE-bench Verified 76.5% · GPQA Diamond 79.2%）
- 高达 52 Token/秒的生成吞吐量，首字延迟 (TTFT) 仅约 1750 毫秒
- 极具竞争力的 API 定价（每百万 Token $10 输入 / $50 输出），原生支持提示词缓存、函数调用与结构化 JSON 输出
Considerations & Trade-offs
- 闭源专有云端模型，不支持下载权重进行离线本地物理隔离部署
- 在大规模高频调用场景下，建议结合提示词缓存 (Prompt Caching) 与批量 API 以控制成本
Token Use & Task Economics
Estimated cost and token consumption to complete one representative agentic task.
| Metric | Fugu Ultra | Claude Fable 5 |
|---|---|---|
Cost per Task Estimated USD to finish one task | $0.28 | $10.97 |
Output Tokens / Task Tokens generated per task | 22K | 72K |
Reasoning Tokens / Task Thinking tokens burned per task | 8K | 37K |
技术规格参数
Architecture, release timeline, capabilities, and modalities.
| Feature | Fugu Ultra | Claude Fable 5 |
|---|---|---|
| Official API Model ID (SDK) | fugu-ultra | claude-fable-5-20260601 |
| API Protocol | Sakana AI REST API (OpenAI-Compatible) | Anthropic Messages API + AWS Bedrock + Vertex AI |
| Release Date | 2026-06 | 2026-06 |
| Knowledge Cutoff | March 2026 | March 2026 |
| Architecture | Evolutionary Multi-Model Merged Transformer | Mythos-Class Dense Reasoning Architecture |
| Parameters | Confidential / Frontier Merged | Confidential |
| Enterprise Rate Limits (RPM / TPM) | 5,000 – 15,000 RPM (Production Tier) · 4,000,000 – 10,000,000 TPM | 4,000 – 10,000 RPM (Tier 4 + Priority Tier) · 4,000,000 – 10,000,000 TPM |
| Long-Context Tier (>128K/200K) | Flat rate ($1.5/1M in) across full 1000K window | Standard rate up to 200K tokens; 2x input/output tier above 200K tokens |
| Prompt Cache Policy & TTL | Cached input at $0.15/1M tokens | $1/1M cached read (90% off); 5-min default & 1-hour extended TTL |
| Computer Use / GUI Agent | API / CLI Tool-Use & DOM Agent ready (no raw pixel desktop click loop) | Supported (Screen coordinate grounding, browser & desktop GUI action loops) |
| MCP & Parallel Tool Calling | Native MCP Client/Server ready · Parallel function calling · 100% Strict JSON Schema | Native MCP Client/Server ready · Parallel function calling · 100% Strict JSON Schema |
| Web Search Grounding & Sandbox | Supported via external Tavily/Exa Search & E2B/Docker Sandbox tools | Native Web Search Citations + Code Execution Tool + MCP Connectors |
| Official Fine-Tuning Support | API Prompt Engineering & RAG Distillation | Custom Model Tuning via Amazon Bedrock & Prompt Distillation |
| Multimodal Token Billing | Text BPE (~4 chars/token) | Text BPE (~4 chars/token) · Vision (~258–765 tokens per image tile) |
| Quantization & Min GPU VRAM | Cloud-Hosted Proprietary (FP8 / BF16 Tensor Core Cluster) · Zero Local VRAM Required (100% Serverless Cloud API) | Cloud-Hosted Proprietary (FP8 / BF16 Tensor Core Cluster) · Zero Local VRAM Required (100% Serverless Cloud API) |
| Safety & Frontier Classification | Sakana AI Production Safety & Alignment Verified | Anthropic RSP ASL-2 / ASL-3 Constitutional Safety Evaluated |
| Enterprise Compliance (ZDR/HIPAA/SOC2) | SOC2 Type II · GDPR · Zero Data Retention (ZDR) & HIPAA BAA Available on Enterprise Tier | SOC2 Type II · GDPR · Zero Data Retention (ZDR) & HIPAA BAA Available on Enterprise Tier |
| Input Modalities | text | text, vision |
| Output Modalities | text | text |
Fugu Ultra vs Claude Fable 5: 常见问题解答 (FAQ)
Which is more intelligent, Fugu Ultra or Claude Fable 5?
Fugu Ultra is more intelligent. It scores 86 compared with 72 on the composite Intelligence Index.
Which model is faster?
Fugu Ultra is faster, generating 75 tokens per second compared with 52 tokens per second.
Which model is cheaper to run?
Fugu Ultra is cheaper at $1.01 per 1M tokens blended, compared with $7.70 (7:2:1 cache:input:output ratio).
Which model has lower latency?
Fugu Ultra has lower latency, with a time to first token of 0.54s compared with 1.75s.
综合结论
综合来看，Fugu Ultra（适合：专为博士级数理逻辑、科学推导与多步思维链 (CoT) 设计的深度推理任务、企业级 RAG 检索增强生成、自动化智能体工作流与高可靠 API 集成。）与 Claude Fable 5（适合：面向复杂企业级推理、软件开发与全场景生产工作流的旗舰前沿任务、企业级 RAG 检索增强生成、自动化智能体工作流与高可靠 API 集成。）各具优势：Fugu Ultra 在综合智能与编程基准测试中表现更强，Fugu Ultra 拥有更快的 Token 生成速度，而 Fugu Ultra 则在规模化 API 部署中具备更高的成本效益。
常见问题 (FAQ)
在编程和深度推理方面，Fugu Ultra 和 Claude Fable 5 哪个更强？+
在综合智能指数与编程基准评测中，Fugu Ultra 表现占优。建议结合上方的 SWE-bench Verified 与 GPQA Diamond 跑分根据具体场景选择。
Fugu Ultra 与 Claude Fable 5 的 API 价格哪个更便宜？+
Fugu Ultra 的 API 价格为输入 $1.50 / 输出 $6.00（每百万 Token），Claude Fable 5 为输入 $10.00 / 输出 $50.00。综合混合成本来看，Fugu Ultra 更具性价比。
哪款模型拥有更大的上下文窗口和更快的响应速度？+
Fugu Ultra 支持 1M Token 上下文窗口，而 Claude Fable 5 支持 1M Token。在流式生成速度方面，Fugu Ultra 的吞吐表现更快。
Understand key metrics like context window, token pricing, reasoning mode, and SWE-bench scores in our comprehensive AI glossary.
Related reading
- Claude Fable 5 vs Claude Sonnet 5：基准测试、API 价格与速度全面对比 (2026)对比
- Claude Fable 5 vs Nex-N2-Mini：基准测试、API 价格与速度全面对比 (2026)对比
- Fugu Ultra vs Qwen3.7 Plus：基准测试、API 价格与速度全面对比 (2026)对比
- Aion-3.0 vs Aion-Mini：基准测试、API 价格与速度全面对比 (2026)对比
- Aion-3.0 vs Gemini 3.6 Flash：基准测试、API 价格与速度全面对比 (2026)对比
- Aion-Mini vs Laguna XS 2.1：基准测试、API 价格与速度全面对比 (2026)对比
Benchmark scores, token prices, and rate limits evolve rapidly as AI labs release model updates. Always verify current specifications and pricing on the official provider documentation.
数据来源: Sakana AI 与 Anthropic 官方 API 文档、模型发布说明及独立基准测试报告 (2026)。