Paper InstrumentsUltramarineFeatherPaper OfficeResearchAn open model for professional workUltramarine 1 Flash is Fable-class on knowledge work spanning finance, law, consulting, and operations. It is also 60x cheaper than Fable.Ultramarine API →Model weights coming soonTechnical report →JobBenchUltramarine 1 Flash62.9Gemini 3.8 Flash60.6DeepSeek V4.1 Flash60.4Claude Fable 557.4GLM-5.3 Flash56.3Kimi K354.3GPT-5.6 Sol45.4PI Knowledge Work (internal)Ultramarine 1 Flash59.3Claude Fable 557.8GPT-5.6 Sol55.1GLM-5.3 Flash52.0Kimi K351.2Gemini 3.8 Flash49.4DeepSeek V4.1 Flash47.9FrontierFinanceUltramarine 1 Flash53.3GLM-5.3 Flash50.7Gemini 3.8 Flash50.0Claude Fable 549.2GPT-5.6 Sol46.8Kimi K346.4DeepSeek V4.1 Flash46.0AutomationBenchDeepSeek V4.1 Flash51.3Ultramarine 1 Flash50.2GLM-5.3 Flash48.8Gemini 3.8 Flash47.3Kimi K346.7Claude Fable 546.2GPT-5.6 Sol45.8DiligenceBenchGemini 3.8 Flash53.1Claude Fable 552.9Ultramarine 1 Flash50.6Kimi K350.4DeepSeek V4.1 Flash50.3GPT-5.6 Sol50.0GLM-5.3 Flash48.2APEX-Agents 1.1Gemini 3.8 Flash64.3Claude Fable 563.6Ultramarine 1 Flash57.0GLM-5.3 Flash52.8GPT-5.6 Sol51.5Kimi K350.6DeepSeek V4.1 Flash39.5Across the six benchmarks, Ultramarine 1 Flash generates 6.5% fewer tokens than its base, GLM-5.3 Flash, taking a shorter path to better work.GLM-5.3 Flash32.7MUltramarine 1 Flash30.5M−6.5%Output tokens generated across all six benchmarks, reasoning included. Where GLM-5.3 Flash’s score is a public one, its tokens come from our own runs closest to that score.Ultramarine 1 Flash sits on the cost-quality frontier at the most attractive price point. At $0.13 per task, it outscores Claude Fable 5 (max) and GPT-5.6 Sol (max), at $8.19 and $2.09 per task, and comes within three points of Claude Opus 5.5 (max) and GPT-6 Sol (max) at $6.96 and $1.26 per task. We measure this on PI Knowledge Work, a carefully built internal benchmark whose scores correlate strongly with the Artificial Analysis Intelligence Index (r = 0.81).Per million tokens on the Ultramarine API, the same as GLM-5.3 Flash.42465054586266$0.01$0.03$0.1$0.3$1$3$10Cost per task (log scale)ScoreUltramarine 1 Flash (max)GLM-5.3 Flash (max)GPT-6 Astra (low)GPT-6 Astra (medium)GPT-6 Astra (max)Claude Opus 5.5 (low)Claude Opus 5.5 (medium)Claude Opus 5.5 (high)Claude Opus 5.5 (max)GPT-6 Sol (low)GPT-6 Sol(medium)GPT-6 Sol (high)GPT-6 Sol (max)Claude Fable 5 (low)Claude Fable 5 (medium)Claude Fable 5 (max)Claude Fable 5.1 (max)GPT-5.6 Sol (low)GPT-5.6 Sol(medium)GPT-5.6 Sol (high)GPT-5.6 Sol (max)Muse Spark 1.3 (low)Muse Spark 1.3 (medium)Muse Spark 1.3 (high)Muse Spark 1.3 (max)GPT-6 Luna (low)GPT-6 Luna (medium)GPT-6 Luna (high)GPT-6 Luna (max)Gemini 3.8 Flash (low)Gemini 3.8 Flash (medium)Gemini 3.8 Flash (high)DeepSeek V4.1 Flash (max)Kimi K3 (max)GLM-5.3 (max)DeepSeek V4 Pro (max)Ling 3.1 Flash (reasoning)Qwen3.8 2.4T (xhigh)Qwen3.8 27B (xhigh)MiniMax M3 (reasoning)PI Knowledge Work is our internal benchmark of professional work in finance, law, consulting, and operations. Its tasks ask the model to deliver finished Excel workbooks, PowerPoint decks, and Word documents, and to work inside live business systems such as ERP and MES.A rubric written for each task scores the result on substance, accuracy, and presentation. Every model works in the same harness, and costs are at each developer's official API pricing. Read our technical report for more on the kind of professional data and evaluations that trained this model.Ultramarine 1 Flash writes clean, simple, straightforward prose and has a strong sense of visual and information hierarchy. It is built to adhere to the same rules required of analysts in professional services.Ultramarine API→Feather→Technical report→GitHub↗Hugging FaceModel weights coming soon