Coding Agents

A coding-specialized model paired post-training with an iterative generate-evaluate-refine loop to exceed the top IOI 2026 human score. The reusable idea is feedback-driven test-time search.

Gemini 3.8 Flash brings multimodal input, tool calling, web search, and default reasoning to coding agents through Vercel. Its temporary 50% discount runs through December 31.

GLM-5.3 is half-price through September 8 via a temporary DigitalOcean-only model ID. Keep the standard ID in durable agent configs if you need fallback after the offer.

Muse Spark 1.3 gives coding agents a 1M-token, multimodal model through Vercel, with a cheaper contributor tier that permits Meta to train on submitted inputs and outputs.

Vercel’s gateway now exposes a pinned Qwen snapshot aimed at larger coding projects and longer agent runs, with a dated model ID that prevents silent upgrades.

An industrial code-model case study frames post-training as maintaining data mixtures under fixed budgets. Raising usable supervision yield improved coding benchmarks without changing the teacher.

Vercel's AI SDK harness now runs fx through the same HarnessAgent API used for other coding agents, reducing the need for another bespoke integration.

AI Gateway can now cap each user's aggregate spend across attributed API keys and app tokens, giving unattended coding agents a hard cost boundary.

As coding agents make implementation easier to copy, builders should spend more judgment on problem choice and preserve claims, evidence, and limits as AI remixes work across product and GTM.

Ironclad argues against optimizing token volume or spend alone: measure reviewed, merged code and fix review and CI bottlenecks before judging coding-agent ROI.

Formal verification gives agent-generated code a machine-checkable contract beyond tests. The hard part shifts upstream: humans must define and validate what correct means.

Figma’s cautious agent adopters expose missing safeguards, not resistance. Use their feedback to improve verification, plans, review size, and the labeling of AI-generated work.

Amazon’s pilot suggests coding-agent gains come from redesigning the workflow: prepare agent-readable context, fast local feedback, scoped specs, and parallel work instead of constant chat.

A study of 8,351 Claude Code plugins finds instruction files and scripts often require coordinated maintenance, treating skills as evolving software rather than static prompts.

OpenAI says it will stop supplying models to Cursor after SpaceX acquired the company, creating a model-availability risk for builders whose workflows depend on Cursor.

Tencent’s Hy4 Preview is now callable through Vercel AI Gateway and selectable in coding agents, adding an open MoE option with a 1M-token context window.

SWE-Prime finds that filtering coding-agent traces by process and segment quality can beat training on every resolved trajectory, reducing noisy imitation from redundant or risky steps.

A multi-GPU kernel benchmark shows coding agents can improve with tools and iteration, but still struggle with communication scheduling, partitioning, and transfer choices.

Mike Krieger describes shifting from stepwise prompting to goal-level delegation, backed by verification, observability, feature flags, and frequent decisions to continue or stop experiments.

Vercel’s AI SDK can now run Cursor behind the same HarnessAgent interface as seven other coding agents, reducing application changes when swapping harnesses.

Agents increasingly choose developer tools, so test whether your docs connect real user pain to your product—not merely whether comparison prompts mention it.

For fast-changing libraries, ship concise agent-facing docs inside the package as well as on the web; agents may inspect installed code before visiting your site.

Qwen 3.8 Flash is now selectable in Vercel AI Gateway and coding agents, with text-and-image input, a 1M-token context window, and responses up to 65k tokens.

GLM 5.3 Flash joins Vercel AI Gateway with text and vision input, a 1M-token context window, function calling, structured output, and streaming.

OpenAI’s Admin plugin lets workspace operators inspect usage and handle membership, permissions, limits, and admin requests from ChatGPT Work or Codex.

AI Gateway offers temporary free routes for MiniMax M3 and M2.7, but the -free model IDs hard-fail after September 6; standard IDs preserve provider fallback at normal rates.

Uber’s agent adoption rests on shared gateways, ready-to-run environments, skills, and a context graph. The operational bottleneck is shifting from code generation to validation and capacity.

GPT-5.6 Sol now costs less across Vercel AI Gateway tiers, with an additional 50% discount through September 18 and no model-ID change for existing agent setups.

DeepSeek V4 Flash Vision adds screenshot, image, and chart input to Vercel AI Gateway while retaining tool use, reasoning, and caching, but its experimental ID signals production risk.

AI4AI-Bench tests whether agents can rewrite training algorithms, not merely tune runs. The best system scored 0.250, with most submissions never changing how the model learns.

A manager can turn fragmented time into agent steering: prepare context, run bounded work overnight, then personally test and review small PRs before involving the team.

Automated code review depends less on a stronger model than on codifying team rules, architecture, incidents, and service contracts into context both humans and agents can inspect.

Vercel Agent can now investigate incidents, edit code, and review PRs inside shared Slack code channels, with approval gates and an audit trail for every action.

Stampli used Codex and ChatGPT Work to cut launch-production hours by 68%, showing how agents can absorb execution work when deadlines and design capacity collide.

Maven’s agent-heavy workflow shortens planning, caps reviewable PR size, and tests nondeterministic features repeatedly. The bottleneck shifts from implementation to decisions and verification.

Replit’s Free Mode uses GPT-5.6 Luna to remove token-cost concerns from initial software creation, lowering friction for experimentation inside its agent workflow.

Nereu argues that coding agents need an engine-native vocabulary, not more generated glue code: tags, reusable systems, and selective scene context turn game intent into reliable actions.

Cursor explains why Git hosting resists naive distribution: DAG walks and delta-packed files punish network round trips, favoring local NVMe replicas kept consistently in sync.

Asana says Codex replaced an outdated test system in two weeks for about $12K, compressing an estimated five-year migration into a focused agent-assisted project.

GLM 5.3 is available through Vercel AI Gateway for coding agents, retaining a 1M-token context window while claiming better long-horizon engineering with fewer output tokens.