可进入对比
档案摘要 / Decision snapshot
先判断能不能用于选型。
有真实任务证据
来自模型卡
边界优先
30 秒结论
可进入选型对比,但仍以案例证据为准。
来自公开模型资料库,并已补厂商/官方/模型家族证据;当前站内暂无可核验 A 类案例。
这页有来源、有风险说明、有公开可核验 A 类案例。
数据可信度 / 证据强度
强
A 类案例
10
AA 评分
25.87
官方来源
1 个补充入口
案例来源
10 条可核验
来自公开资料与模型数据库 + Artificial Analysis + 1 个厂商/官方/模型家族证据入口;AA 不是唯一事实来源。
Evidence Distribution
案例证据分布
Solar Pro 3 当前关联 10 条 A 类案例;这里按任务、来源和复核时间观察证据结构。
Evidence Snapshot
证据快照覆盖
Solar Pro 3 的 10 条 A 类案例中,已有 10 条完成本地证据快照;0 条仍在快照队列。
可用 manifest 复核
等待落盘
需补抓
需人工复核
证据完整度
5 / 5
Feishu Bitable
1 个入口
10 条
已绑定
已标注
能力边界
用标签表达信号,不用标签替代证据。
通用/待核验
暂无数据
待核验
平台/待核验
中
| 厂商 | Upstage / Solar |
|---|---|
| 厂商页 | Upstage / Solar |
| 发布时间 | 2026-04-06 |
| 模型 ID | solar-pro-3 |
| 上下文 | 暂无数据 |
| 输出 | 暂无数据 |
| 模态 | 文本 / 多模态待核验 |
| 推理 | 是 / 推理模型或 thinking 模式 |
| 价格 | 官方未披露 / 暂无数据 |
| 可用平台 | Upstage / Solar |
| AA 评分 | 25.87 |
| 基础模型发布时间 | Artificial Analysis 未披露 |
| 资料来源 | 公开资料、厂商信息和案例库 |
| A 类案例 | 10 |
谱系位置
它在厂商路线中的位置
此页将 Solar Pro 3 放入 Upstage / Solar 的当前 Atlas 路线中。若同厂前后代资料不足,先保留为路线追踪入口,不脑补谱系关系。
2026-04-06
Solar Pro 3 资料状态
来自公开资料与模型数据库 + Artificial Analysis + 1 个厂商/官方/模型家族证据入口;AA 不是唯一事实来源。
公开档案
模型状态
Solar Pro 3 当前标记为已有真实案例。
案例库
公开案例复核
已有 10 条可核验 A 类案例,达到完整补齐线
案例时间线
最近进入档案的 A 类案例
用采集时间呈现证据进入 Atlas 的顺序,方便复核来源新鲜度。
2026-06-27T08:34:23Z
SolarNote: Solar Pro 3 multi-agent math error-note generator
junjunjunbong / education
2026-06-27T08:34:23Z
Tax Preparation Assistant using Solar Pro 3 via OpenRouter
Samson-U / finance_tax
2026-06-27T08:34:23Z
Bookitty Kitty accounting chatbot with Solar Pro 3 first in fallback chain
AndyNope / Bookitty / finance_accounting
2026-06-27T08:34:23Z
G-ONE personal task automation chatbot exposes Solar Pro 3 as a selectable OpenRouter model
ARUN-L-KUMAR / G-ONE / productivity_agent
2026-06-27T08:34:23Z
Multi-Agent Interview Coach configured with Solar Pro 3
EmilBBbbbbbb / hr_recruiting
适合场景
- - 作为完整模型索引和厂商路线追踪入口
- - 推理/agentic workflow 候选
不适合
- - 缺少真实案例时不要包装成推荐
- - 缺字段时不要脑补价格、性能或上下文
真实案例
可核验 A 类案例精选
当前展示 10 条代表案例;排序优先看模型卡精选、证据可信度、快照状态和公开仓库/技术任务信号。案例总数仍为 10。
Representative Evidence
代表案例排序
当前第一代表案例是“Bookitty Kitty accounting chatbot with Solar Pro 3 first in fallback chain”。它会优先出现在本页,是因为该案例同时进入模型卡精选、具备 100/100 的证据可信度,并带有 已快照证据。
模型卡精选优先
A+ 完整链路
archive/evidence/solar-pro-3_andynope_bookitty_20260627/manifest.json
Coding / repo / agent
AndyNope / Bookitty 使用 Solar Pro 3 处理知识检索和问答
AndyNope / Bookitty · Solar Pro 3
A+ 完整链路 · 代码仓库证据
AndyNope / Bookitty 公开的知识库与检索问答案例,来源为 公开代码库,复核于 2026-06-27T08:34:23Z。 Model Atlas 将它标记为 A 类证据,因为它同时具备具体使用者、具体任务、公开原始证据和可访问产物。
任务:这是一个围绕知识检索和问答的真实任务,公开材料可以回溯到具体使用者和具体产物。 原始资料写作:Bookitty is a Swiss SME accounting web app with Kitty, an AI chatbot that answers accounting, VAT, and app-usage questions, can suggest bookkeeping entries, and can trigger interface actions such as navigation highlight…
公开产物:公开材料提供公开代码、README 或项目配置,可用于核验任务结果、项目形态和模型绑定关系。 原始资料写作:The public README states that the app is deployed at bookitty.bidebliss.com and that Kitty provides an offline-first knowledge base plus an OpenRouter fallback chain for questions outside the knowledge base; api/chat.ph…
模型作用:Solar Pro 3 在该案例中承担知识检索和问答相关的生成、分析、编排或实现角色。 原始资料写作:api/chat.php places upstage/solar-pro-3:free as the first model in the OpenRouter fallback chain, so Solar Pro 3 is the first external LLM attempted for out-of-knowledge-base accounting-chatbot responses before other fr…
A 类理由:有具体使用者、具体任务、公开原始证据和可访问产物,可公开核验。
风险边界:当前判断基于公开材料;若产物下线、仓库变更或模型参与比例仅来自作者自述,需要在引用前重新复核。 原始资料写作:Solar Pro 3 is used as first fallback rather than the only model; the artifact is a public product page plus repo evidence, but no per-model live traffic metrics are public.
原始记录:Bookitty Kitty accounting chatbot with Solar Pro 3 first in fallback chain
ARUN-L-KUMAR / G-ONE 使用 Solar Pro 3 处理软件工程任务执行
ARUN-L-KUMAR / G-ONE · Solar Pro 3
A+ 完整链路 · 代码仓库证据
ARUN-L-KUMAR / G-ONE 公开的代码代理与软件工程案例,来源为 公开代码库,复核于 2026-06-27T08:34:23Z。 Model Atlas 将它标记为 A 类证据,因为它同时具备具体使用者、具体任务、公开原始证据和可访问产物。
任务:这是一个围绕软件工程任务执行的真实任务,公开材料可以回溯到具体使用者和具体产物。 原始资料写作:G-ONE is a full-stack personal productivity platform that combines calendar, tasks, Gmail, contacts, maps, sheets, and notes context to answer user questions, detect conflicts, prioritize work, and produce optimized dai…
公开产物:公开材料提供公开代码、README 或项目配置,可用于核验任务结果、项目形态和模型绑定关系。 原始资料写作:The repository documents manual and live planning modes, multi-agent orchestration, persistent AI plans, and a chatbot endpoint that returns replies with context sources, agents used, model source, latency, and fallback…
模型作用:Solar Pro 3 在该案例中承担软件工程任务执行相关的生成、分析、编排或实现角色。 原始资料写作:chatbot_router.py exposes Solar Pro 3 as the OpenRouter option openrouter-alt2, routes explicit user selection of that model into the LLM fallback list, maps successful calls back to 'OpenRouter (Solar Pro 3)', and incl…
A 类理由:有具体使用者、具体任务、公开原始证据和可访问产物,可公开核验。
风险边界:当前判断基于公开材料;若产物下线、仓库变更或模型参与比例仅来自作者自述,需要在引用前重新复核。 原始资料写作:Solar Pro 3 is one selectable/fallback model in a multi-provider system, not the sole model. Evidence is public source and README documentation, with no public deployment metrics.
原始记录:G-ONE personal task automation chatbot exposes Solar Pro 3 as a selectable OpenRouter model
EmilBBbbbbbb 使用 Solar Pro 3 处理软件工程任务执行
EmilBBbbbbbb · Solar Pro 3
A+ 完整链路 · 代码仓库证据
EmilBBbbbbbb 公开的代码代理与软件工程案例,来源为 公开代码库,复核于 2026-06-27T08:34:23Z。 Model Atlas 将它标记为 A 类证据,因为它同时具备具体使用者、具体任务、公开原始证据和可访问产物。
任务:这是一个围绕软件工程任务执行的真实任务,公开材料可以回溯到具体使用者和具体产物。 原始资料写作:A multi-agent technical interview simulator in which interviewer, observer, evaluator, and feedback-generator agents conduct an interview, analyze candidate responses, adapt question difficulty, evaluate correctness, an…
公开产物:公开材料提供公开代码、README 或项目配置,可用于核验任务结果、项目形态和模型绑定关系。 原始资料写作:The README describes a runnable CLI workflow that asks for candidate profile details, conducts the interview turn by turn, records internal agent thoughts, supports finish/status commands, and stores interview logs as J…
模型作用:Solar Pro 3 在该案例中承担软件工程任务执行相关的生成、分析、编排或实现角色。 原始资料写作:The README's environment setup selects LLM_PROVIDER=openrouter and sets OPENROUTER_MODEL=upstage/solar-pro-3:free, binding Solar Pro 3 to the agent LLM calls used for interview question generation, response analysis, ev…
A 类理由:有具体使用者、具体任务、公开原始证据和可访问产物,可公开核验。
风险边界:当前判断基于公开材料;若产物下线、仓库变更或模型参与比例仅来自作者自述,需要在引用前重新复核。 原始资料写作:Public evidence is from the repository README and appears to be a prototype/demo rather than a production customer deployment; still includes a concrete runnable artifact and exact Solar Pro 3 model binding.
原始记录:Multi-Agent Interview Coach configured with Solar Pro 3
junjunjunbong 使用 Solar Pro 3 处理软件工程任务执行
junjunjunbong · Solar Pro 3
A+ 完整链路 · 代码仓库证据
junjunjunbong 公开的代码代理与软件工程案例,来源为 公开代码库,复核于 2026-06-27T08:34:23Z。 Model Atlas 将它标记为 A 类证据,因为它同时具备具体使用者、具体任务、公开原始证据和可访问产物。
任务:这是一个围绕软件工程任务执行的真实任务,公开材料可以回溯到具体使用者和具体产物。 原始资料写作:A Streamlit application for students and tutors that accepts a photo of a wrong math problem, parses the problem, classifies the concept and difficulty, generates five similar problems, writes step-by-step solutions, an…
公开产物:公开材料提供公开代码、README 或项目配置,可用于核验任务结果、项目形态和模型绑定关系。 原始资料写作:The public repository documents a working multi-agent pipeline whose final output is an error note containing five similar problems, detailed solutions, formulas, learning advice, and next-study recommendations.
模型作用:Solar Pro 3 在该案例中承担软件工程任务执行相关的生成、分析、编排或实现角色。 原始资料写作:Solar Pro 3 is explicitly named in the repository as the LLM used by ConceptAgent, ProblemAgent, SolutionAgent, and NoteAgent for concept classification, similar-problem generation, solution writing, and error-note summ…
A 类理由:有具体使用者、具体任务、公开原始证据和可访问产物,可公开核验。
风险边界:当前判断基于公开材料;若产物下线、仓库变更或模型参与比例仅来自作者自述,需要在引用前重新复核。 原始资料写作:Open-source project evidence; no independent production usage metrics were found. The README and code bind the artifact to Solar Pro 3, while OCR/extraction stages use other Upstage APIs.
原始记录:SolarNote: Solar Pro 3 multi-agent math error-note generator
Samson-U 使用 Solar Pro 3 处理软件工程任务执行
Samson-U · Solar Pro 3
A+ 完整链路 · 代码仓库证据
Samson-U 公开的代码代理与软件工程案例,来源为 公开代码库,复核于 2026-06-27T08:34:23Z。 Model Atlas 将它标记为 A 类证据,因为它同时具备具体使用者、具体任务、公开原始证据和可访问产物。
任务:这是一个围绕软件工程任务执行的真实任务,公开材料可以回溯到具体使用者和具体产物。 原始资料写作:A Streamlit tax-filing chatbot that conversationally collects financial inputs, applies compressed tax-code and deduction-guide context, calculates simplified tax liability, and provides personalized tax-saving suggesti…
公开产物:公开材料提供公开代码、README 或项目配置,可用于核验任务结果、项目形态和模型绑定关系。 原始资料写作:The repository provides a runnable Streamlit app with chat logic, API clients, settings, and environment instructions; its README describes the assistant's output as personalized tax preparation guidance and tax-saving …
模型作用:Solar Pro 3 在该案例中承担软件工程任务执行相关的生成、分析、编排或实现角色。 原始资料写作:The README and settings.py bind the OpenRouter model configuration to upstage/solar-pro-3:free by default, making Solar Pro 3 the LLM that interprets user financial inputs and produces conversational tax guidance.
A 类理由:有具体使用者、具体任务、公开原始证据和可访问产物,可公开核验。
风险边界:当前判断基于公开材料;若产物下线、仓库变更或模型参与比例仅来自作者自述,需要在引用前重新复核。 原始资料写作:The project is explicitly described as a learning/prototyping demo and not legal or tax advice; public evidence is repo-level rather than a deployed customer story.
原始记录:Tax Preparation Assistant using Solar Pro 3 via OpenRouter
gem-squared 使用 Solar Pro 3 处理软件工程任务执行
gem-squared · Solar Pro 3
A+ 完整链路 · 代码仓库证据
gem-squared 公开的代码代理与软件工程案例,来源为 公开代码库,复核于 2026-06-27T08:18:08Z。 Model Atlas 将它标记为 A 类证据,因为它同时具备具体使用者、具体任务、公开原始证据和可访问产物。
任务:这是一个围绕软件工程任务执行的真实任务,公开材料可以回溯到具体使用者和具体产物。 原始资料写作:Implement a runtime judgment verification layer for Korean health-insurance claim documents, combining Upstage Document AI extraction with Solar Pro 3 pre-check, judgment and evidence-audit stages.
公开产物:公开材料提供公开代码、README 或项目配置,可用于核验任务结果、项目形态和模型绑定关系。 原始资料写作:The public demo and repository include three preloaded scenarios: a clean claim, a trap claim where verification catches an unsupported rider certificate claim, and a malicious-injection claim blocked before LLM process…
模型作用:Solar Pro 3 在该案例中承担软件工程任务执行相关的生成、分析、编排或实现角色。 原始资料写作:Solar Pro 3 performs L1 precondition validation, the core Korean-language adjudication over extracted facts, and L2 epistemic auditing that traces cited evidence references back to source anchors.
A 类理由:有具体使用者、具体任务、公开原始证据和可访问产物,可公开核验。
风险边界:当前判断基于公开材料;若产物下线、仓库变更或模型参与比例仅来自作者自述,需要在引用前重新复核。 原始资料写作:Evidence is self-declared by the public repository and demo; it is a demo governance layer, not a named insurer production case.
原始记录:Solar Trust Gate uses Solar Pro 3 to adjudicate and audit Korean insurance claims
Hongduk 使用 Solar Pro 3 处理知识检索和问答
Hongduk · Solar Pro 3
A+ 完整链路 · 代码仓库证据
Hongduk 公开的知识库与检索问答案例,来源为 公开代码库,复核于 2026-06-27T08:18:08Z。 Model Atlas 将它标记为 A 类证据,因为它同时具备具体使用者、具体任务、公开原始证据和可访问产物。
任务:这是一个围绕知识检索和问答的真实任务,公开材料可以回溯到具体使用者和具体产物。 原始资料写作:Create a personal Korean exam-study tutor for a middle-school student by parsing scanned textbook PDFs with Upstage Document Parse and loading each unit directly into Solar Pro 3's long context for five study modes.
公开产物:公开材料提供公开代码、README 或项目配置,可用于核验任务结果、项目形态和模型绑定关系。 原始资料写作:The app offers unit overview, key concept review, step-by-step interactive tutoring, quiz generation with grading/explanations and mock exams through a Streamlit interface.
模型作用:Solar Pro 3 在该案例中承担知识检索和问答相关的生成、分析、编排或实现角色。 原始资料写作:Solar Pro 3 is the default tutoring LLM; its 128K context window is used instead of RAG to keep an entire unit in context and generate Korean study explanations, questions, grading and feedback.
A 类理由:有具体使用者、具体任务、公开原始证据和可访问产物,可公开核验。
风险边界:当前判断基于公开材料;若产物下线、仓库变更或模型参与比例仅来自作者自述,需要在引用前重新复核。 原始资料写作:Evidence is self-declared in a public GitHub README; textbook files are excluded for copyright, but the repository describes the workflow and runnable app structure.
原始记录:Korean middle-school personal tutor app uses Solar Pro 3 long context over textbook units
Hyeongseob91 使用 Solar Pro 3 处理软件工程任务执行
Hyeongseob91 · Solar Pro 3
A+ 完整链路 · 代码仓库证据
Hyeongseob91 公开的代码代理与软件工程案例,来源为 公开代码库,复核于 2026-06-27T08:18:08Z。 Model Atlas 将它标记为 A 类证据,因为它同时具备具体使用者、具体任务、公开原始证据和可访问产物。
任务:这是一个围绕软件工程任务执行的真实任务,公开材料可以回溯到具体使用者和具体产物。 原始资料写作:Build a Streamlit dashboard that crawls Moltbook posts and uses Upstage Solar Pro 3 to classify topics, writing patterns, agent personas, sentiment, trends and meme signals in AI-agent community discourse.
公开产物:公开材料提供公开代码、README 或项目配置,可用于核验任务结果、项目形态和模型绑定关系。 原始资料写作:The repository provides a runnable dashboard with real-time crawling, background analysis, SQLite caching, four analysis tabs, export features and an analysis schema for Solar Pro 3 outputs.
模型作用:Solar Pro 3 在该案例中承担软件工程任务执行相关的生成、分析、编排或实现角色。 原始资料写作:Solar Pro 3 is the LLM analysis engine for comprehensive post analysis, producing structured labels for topic, persona, sentiment and trend fields that drive the dashboard views.
A 类理由:有具体使用者、具体任务、公开原始证据和可访问产物,可公开核验。
风险边界:当前判断基于公开材料;若产物下线、仓库变更或模型参与比例仅来自作者自述,需要在引用前重新复核。 原始资料写作:Evidence is self-declared in a public GitHub README; it is an open-source project rather than independently verified production deployment.
原始记录:Agent Genome Watcher analyzes Moltbook AI-agent community discourse with Solar Pro 3
Mixup-Agent / solsol team 使用 Solar Pro 3 处理软件工程任务执行
Mixup-Agent / solsol team · Solar Pro 3
A+ 完整链路 · 代码仓库证据
Mixup-Agent / solsol team 公开的代码代理与软件工程案例,来源为 公开代码库,复核于 2026-06-27T08:18:08Z。 Model Atlas 将它标记为 A 类证据,因为它同时具备具体使用者、具体任务、公开原始证据和可访问产物。
任务:这是一个围绕软件工程任务执行的真实任务,公开材料可以回溯到具体使用者和具体产物。 原始资料写作:Build a web-based voice mock-interview service that analyzes resumes, cover letters, portfolios, company names, job postings and company news, then runs multiple interview agents tailored to a target employer's intervie…
公开产物:公开材料提供公开代码、README 或项目配置,可用于核验任务结果、项目形态和模型绑定关系。 原始资料写作:The repository describes a LangGraph multi-agent service with company-style inference, resume/trend/stress interviewers, answer-quality gating, STT/TTS, source-tracked trend questions and final feedback reports.
模型作用:Solar Pro 3 在该案例中承担软件工程任务执行相关的生成、分析、编排或实现角色。 原始资料写作:Solar Pro3 is used for the meta router and judge at high reasoning effort, and for resume, trend and stress interview agents at low reasoning effort; prompt caching and Solar Mini are used for cost/speed tradeoffs.
A 类理由:有具体使用者、具体任务、公开原始证据和可访问产物,可公开核验。
风险边界:当前判断基于公开材料;若产物下线、仓库变更或模型参与比例仅来自作者自述,需要在引用前重新复核。 原始资料写作:Evidence is self-declared in a hackathon/open-source GitHub README; it demonstrates a concrete app but not independently verified commercial production usage.
原始记录:solsol builds a voice-based Korean mock interview service with Solar Pro3 multi-agent routing
Sioonn 使用 Solar Pro 3 处理软件工程任务执行
Sioonn · Solar Pro 3
A+ 完整链路 · 代码仓库证据
Sioonn 公开的代码代理与软件工程案例,来源为 公开代码库,复核于 2026-06-27T08:18:08Z。 Model Atlas 将它标记为 A 类证据,因为它同时具备具体使用者、具体任务、公开原始证据和可访问产物。
任务:这是一个围绕软件工程任务执行的真实任务,公开材料可以回溯到具体使用者和具体产物。 原始资料写作:Improve Korean natural-language fashion image search by generating user-style Korean search queries from product captions and using them to align a SigLIP 2 text tower with fashion product images.
公开产物:公开材料提供公开代码、README 或项目配置,可用于核验任务结果、项目形态和模型绑定关系。 原始资料写作:The project reports a lightweight retrieval pipeline where MRR improves from 0.571 zero-shot SigLIP 2 to 0.760 after text-only LoRA, BM25 fusion and query-bank prior correction on human gold queries.
模型作用:Solar Pro 3 在该案例中承担软件工程任务执行相关的生成、分析、编排或实现角色。 原始资料写作:Solar-pro3 is used as the external API model for training-query generation, creating three Korean user-style synthetic queries per product caption for a 7,023-query bank used in LoRA alignment and prior correction.
A 类理由:有具体使用者、具体任务、公开原始证据和可访问产物,可公开核验。
风险边界:当前判断基于公开材料;若产物下线、仓库变更或模型参与比例仅来自作者自述,需要在引用前重新复核。 原始资料写作:Evidence is self-declared in a public research-code README; the reported metrics are project results and should be treated as repository-authored claims.
原始记录:K-fashion text-to-image retrieval project uses Solar Pro3 to generate Korean user-style synthetic queries
数据缺口
只基于现有字段判断缺口;缺失项不会被猜测填充。
- - 补充方向:核验 价格 的官方/API/案例来源。
- - 补充方向:核验 上下文 的官方/API/案例来源。
- - 补充方向:核验 输出 的官方/API/案例来源。
下一步验证建议
- - 继续保留 A 类案例的原始证据、产物页和版本快照。
- - 补官方发布/API/System Card 来源;若官方未披露,继续标注“官方未披露”。
- - 把 benchmark、教程、发布文保留为背景资料,不提升为真实案例。