Ling-3.0-flash-Fin
Ling-3.0-flash-Fin by InclusionAI is the first finance-enhanced Ling model, adding source-grounded search, valuation modeling, and spreadsheet workflows.
Model Overview
Capabilities, design details, and architectural traits
Ling-3.0-flash-Fin - First Finance-Enhanced Model in the Ant Ling Family
Ling-3.0-flash-Fin is the first finance-enhanced release from Ant Group's InclusionAI team. It extends the Ling-3.0-flash base model through continued training on high-quality financial data, built together with leading financial institutions and domain experts. The goal is a single model that can run complete investment-research workflows instead of isolated tasks.
What Sets It Apart
| Trait | Description |
|---|---|
| End-to-end financial research | Connects information retrieval, evidence review, calculation, modeling, and report preparation as one pipeline rather than separate steps. |
| Source-grounded financial search | Prioritizes authoritative sources to give accurate, complete, and traceable answers. The FinFIRST dataset is open-sourced alongside the model for transparent evaluation of these capabilities. |
| Multi-document financial reasoning | Reconciles reporting periods, definitions, assumptions, and conflicting figures across annual reports, earnings releases, regulatory filings, and research materials. |
| Valuation and spreadsheet workflows | Understands formulas, actual-versus-estimate updates, cross-sheet dependencies, balance checks, scenario analysis, and editable financial-model delivery. |
| Research-ready outputs | Organizes facts, analysis, judgments, and charts into clear, reviewable materials for further editing and professional review. |
| Efficient long-horizon design | Keeps the Ling-3.0-flash architecture with 124B total and 5.1B activated parameters plus a 256K context window, pairing financial expertise with efficient inference for long-horizon agent workflows. |
Built on the Ling-3.0-flash Backbone
The model shares its architecture with Ling-3.0-flash, so it runs on the same SGLang and vLLM setups. Thinking mode is enabled by default, and the model card recommends specific sampling settings for general inference. It is released in BF16 with FP8, INT4, and FP4 quantized variants.
As a finance-focused release, the model card notes that key assumptions, valuation results, and investment conclusions still require professional review and do not constitute investment advice.
Benchmark Performance
Independent evaluations · Artificial Analysis
Accuracy & Capability Details
Compare Models Side-by-Side
Evaluate specifications, pricing, and independent benchmark indices
| Model Details | |||
|---|---|---|---|
| General Info | |||
| Provider | InclusionAI | Anthropic | Anthropic |
| Release Date | September 11, 2026 | September 22, 2026 | September 28, 2026 |
| Knowledge Cutoff | - | - | Jun 2026 |
| Context & Limits | |||
| Context Window | 256K | 1M Best Context Window | 1M Best Context Window |
| Pricing (per 1M tokens) | |||
| Input Pricing | $0.07 Best Input Pricing | $4 | $2 |
| Output Pricing | $0.22 Best Output Pricing | $20 | $10 |
| Modalities | |||
| Inputs | text | textimagefile | textimagefile |
| Outputs | text | text | text |
| Benchmarks (0-100) | |||
| Intelligence Index | 22.6 | 57.6 Best Intelligence Index | 56.0 |
| Coding Index | 55.6 | - | - |
| Agentic Index | 27.9 | - | - |
Humanity's Last Exam
Extremely difficult logical reasoning and knowledge.
Long Context Reasoning
Logical reasoning over long context windows.
SciCode Benchmark
Scientific coding and mathematical modeling.
Independent evaluation data provided by Artificial Analysis. To view the latest benchmarks and full details, visit their official site.
Explore more from InclusionAI
Other models by InclusionAI
Ling 3.1 Flash
Ling-3.0-flash-VL
Ling 3.0 Flash
Ling 3.0 Tiny
Top AI Models
Leading alternatives by intelligence score
Claude Opus 5.5
Muse Spark 1.3
GPT-6 Astra
