{"id":570783,"date":"2026-03-18T18:27:34","date_gmt":"2026-03-18T18:27:34","guid":{"rendered":"https:\/\/Blockchain.News\/news\/together-ai-fine-tuning-tool-calling-reasoning-vision"},"modified":"2026-03-18T18:27:34","modified_gmt":"2026-03-18T18:27:34","slug":"together-ai-upgrades-fine-tuning-platform-with-vision-and-reasoning-support","status":"publish","type":"post","link":"https:\/\/e-bitco.in\/index.php\/2026\/03\/18\/together-ai-upgrades-fine-tuning-platform-with-vision-and-reasoning-support\/","title":{"rendered":"Together AI Upgrades Fine-Tuning Platform With Vision and Reasoning Support"},"content":{"rendered":"
\n

Joerg Hiller<\/a> Mar 18, 2026 18:27<\/span> <\/p>\n

Together AI adds tool calling, reasoning traces, and vision-language fine-tuning to its platform, with 6x throughput gains for 100B+ parameter models.<\/p>\n

\"Together <\/a> <\/figure>\n

Together AI rolled out a major expansion to its fine-tuning service on March 18, adding native support for tool calling, reasoning traces, and vision-language models\u2014capabilities that address persistent pain points for teams building production AI systems.<\/p>\n

The update arrives as the company reportedly negotiates a funding round that would value it at $7.5 billion, more than doubling its $3.3 billion valuation from its February 2025 Series B.<\/p>\n

What’s Actually New<\/h2>\n

The platform now handles three categories of fine-tuning that previously required fragmented workarounds:<\/p>\n

Tool calling<\/strong> gets end-to-end support using OpenAI-compatible schemas. The system validates that every tool call in training data matches declared functions before training begins\u2014a safeguard against the hallucinated parameters and schema mismatches that plague agentic workflows.<\/p>\n

Reasoning fine-tuning<\/strong> allows teams to train models on domain-specific thinking traces using a dedicated reasoning_content field. This matters because reasoning formats vary wildly across model families, making consistent training difficult without standardization.<\/p>\n

Vision-language fine-tuning<\/strong> supports hybrid datasets mixing image-text and text-only examples. By default, the vision encoder stays frozen while language layers update, though teams can enable joint training when visual pattern recognition needs improvement.<\/p>\n

Infrastructure Upgrades<\/h2>\n

Beyond new capabilities, Together AI claims significant performance gains from optimizing its training stack for mixture-of-experts architectures. The company integrated SonicMoE kernels that overlap memory operations with computation, plus custom CUDA kernels for loss computation.<\/p>\n

Results vary by model size: smaller models see roughly 2x throughput improvements, while larger architectures like Kimi-K2 hit 6x gains. The platform now handles datasets up to 100GB and models exceeding 100 billion parameters.<\/p>\n

New models available for fine-tuning include Qwen 3.5 variants (up to 397B parameters), Kimi K2 and K2.5, and GLM-4.6 and 4.7.<\/p>\n

Practical Additions<\/h2>\n

The update includes cost estimation before job execution and live progress tracking with dynamic completion estimates\u2014features that sound basic but prevent the budget surprises that make experimentation risky.<\/p>\n

XY.AI Labs, cited by Together AI as a customer example, reported moving from weekly to daily iteration cycles while cutting costs 2-3x and improving accuracy from 77% to 87% using the platform’s fine-tuning and deployment APIs.<\/p>\n

Market Context<\/h2>\n

The timing aligns with a surge in AI infrastructure spending. Startup funding in the AI sector hit $220 billion in the first two months of 2026, per recent reports, with much of that capital flowing toward training and inference infrastructure.<\/p>\n

Together AI positions itself as an alternative to building in-house AI infrastructure, offering access to over 200 open-source models through its platform. The company’s pitch\u2014removing infrastructure complexity so teams can focus on product development\u2014now extends to increasingly sophisticated post-training workflows that were previously the domain of well-resourced research labs.<\/p>\n

Image source: Shutterstock<\/i><\/span> Source<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"

Joerg Hiller Mar 18, 2026 18:27 Together AI adds tool calling, reasoning traces, and vision-language fine-tuning to its platform, with 6x throughput gains for 100B+ parameter models. Together AI rolled out a major expansion to its fine-tuning service on March 18, adding native support for tool calling, reasoning traces, and vision-language models\u2014capabilities that address persistent […]<\/p>\n","protected":false},"author":1,"featured_media":570784,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[12],"tags":[20916,20460,22004,2572,25,19527],"class_list":{"0":"post-570783","1":"post","2":"type-post","3":"status-publish","4":"format-standard","5":"has-post-thumbnail","7":"category-blockchain","8":"tag-ai-infrastructure","9":"tag-enterprise-ai","10":"tag-fine-tuning","11":"tag-machine-learning","12":"tag-news","13":"tag-together-ai"},"_links":{"self":[{"href":"https:\/\/e-bitco.in\/index.php\/wp-json\/wp\/v2\/posts\/570783","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/e-bitco.in\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/e-bitco.in\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/e-bitco.in\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/e-bitco.in\/index.php\/wp-json\/wp\/v2\/comments?post=570783"}],"version-history":[{"count":0,"href":"https:\/\/e-bitco.in\/index.php\/wp-json\/wp\/v2\/posts\/570783\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/e-bitco.in\/index.php\/wp-json\/wp\/v2\/media\/570784"}],"wp:attachment":[{"href":"https:\/\/e-bitco.in\/index.php\/wp-json\/wp\/v2\/media?parent=570783"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/e-bitco.in\/index.php\/wp-json\/wp\/v2\/categories?post=570783"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/e-bitco.in\/index.php\/wp-json\/wp\/v2\/tags?post=570783"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}