model release

Upstage Releases Solar Pro 4 With 524K Token Context Window at $0.03/M Input Pricing

TL;DR

Upstage has released Solar Pro 4, a large language model with a 524K token context window aimed at agentic workflows, document processing, and coding. The model is priced at $0.03 per million input tokens and $0.12 per million output tokens, and is available now via OpenRouter.

2 min read
0

Solar Pro 4 — Quick Specs

Context window524K tokens
Input$0.03/1M tokens
Output$0.12/1M tokens

Upstage has released Solar Pro 4, a new large language model with a 524,000 token context window, according to a listing on OpenRouter where the model is now available as upstage/solar-pro4.

The model is positioned by Upstage for agentic workflows, office productivity tasks, document-intensive work, and coding. It operates in text-to-text mode only, with no stated support for image, audio, or video inputs.

Pricing and Access

Solar Pro 4 is priced at $0.03 per million input tokens and $0.12 per million output tokens through OpenRouter's API. That places it among the lower-cost options in its context-window class, though independent benchmark results comparing it to competing models have not yet been published.

The 524K token context window is large enough to process lengthy documents, multi-file codebases, or extended multi-turn agentic sessions in a single call, which aligns with Upstage's stated target use cases of document processing and agentic workflows.

What We Don't Know

Upstage has not disclosed the model's parameter count, training data cutoff date, or architecture details in the material accompanying this release. No independent benchmark scores (MMLU, HumanEval, or similar standard evaluations) have been published alongside the launch, so claims about the model's suitability for coding and office productivity tasks remain Upstage's own characterization rather than independently verified performance data.

What This Means

Solar Pro 4 enters a market where long-context, low-cost models are increasingly common, and Upstage's pricing—$0.03/$0.12 per million tokens—undercuts many rivals offering comparable context lengths. The 524K token window is notable because it exceeds the context windows offered by several widely used commercial models, making it a candidate for teams processing large document sets or running extended agent sessions without chunking content.

However, without published benchmark results, buyers evaluating Solar Pro 4 for coding or reasoning-heavy tasks will need to run their own evaluations before committing production workloads to the model. Upstage, a South Korean AI company known for its Solar model family, appears to be competing primarily on context length and price rather than on demonstrated benchmark leadership at this stage. As with any newly released model lacking third-party verification, enterprises should treat throughput, latency, and accuracy claims as provisional until independent testing confirms them.

Related Articles

model release

Alibaba Unveils Qwen3.8-Max, a 2.4T-Parameter Open-Weight Model for Coding and Agentic Work

Alibaba announced Qwen3.8-Max, a 2.4T-parameter flagship model targeting coding and long-horizon agentic work, with open weights promised for next week alongside Qwen3.8-27B. The model posted strong third-party benchmark results, ranking #4 in Frontend Code Arena and matching Claude Opus 4.7 on the Vals Index at roughly 2.3x lower cost.

model release

Meta Releases Muse Glimmer, a 30B Open-Weights Model Under Apache 2.0 License

Meta has released Muse Glimmer, a 30 billion parameter open-weights model licensed under Apache 2.0. The model targets agentic task completion, tool use, and multi-step reasoning, with benchmark results claimed on DeepSearch QA, MCP-Atlas, τ-Bench, and SWE-Bench.

model release

Unsloth Releases GGUF Quantizations of Meta's Muse Glimmer 30B Agentic Model

Unsloth has published GGUF quantizations of Muse Glimmer-30B, a dense 29.6B-parameter causal transformer with a dedicated perception encoder, attributed to Meta Superintelligence Lab in the model card. The model targets autonomous agentic tasks on consumer hardware with a 131,072-token context window and 4-bit quantization under 20GB.

model release

Liquid AI Releases LFM2.5-2.6B, a 2.6B-Parameter Agentic Model with 128K Context for On-Device Use

Liquid AI has released LFM2.5-2.6B, a 2.6B-parameter model trained on 34 trillion tokens with a 128K context window, built for on-device agentic workloads. The company claims it is competitive with models four times its size on tool use and instruction following.

Comments

Loading...