VFF - The signal in the noise
VFF - The signal in the noise
NewsTrending

Tech Firms Build Custom Coding Agents to Cut AI Costs

The Information
Read original
Share
Tech Firms Build Custom Coding Agents to Cut AI Costs

Coinbase, Shopify, and Ramp have built proprietary AI coding agents for internal use to reduce reliance on expensive third-party solutions from Anthropic and OpenAI. Coinbase's tool, called Forge, launched to all engineers in April 2026 and is accessible via Slack, GitHub, and a web interface. The move reflects a broader trend among well-resourced tech firms to develop custom alternatives rather than pay for commercial coding AI services.

TL;DR

  • Coinbase launched Forge, an internal AI coding agent, to all engineers in April 2026
  • Shopify and Ramp have also built proprietary coding agents for employee use
  • Internal tools reduce costs compared to commercial offerings from Anthropic and OpenAI
  • Forge usage is growing and has delivered measurable cost savings, though specifics were not disclosed

Why It Matters

Large tech firms are increasingly building custom AI coding tools rather than relying on commercial vendors, signaling both the maturity of AI agent development and the cost pressures of widespread adoption. This trend could reshape the market for coding AI by fragmenting demand away from vendors like Anthropic and OpenAI toward internal solutions.

Business Impact

For enterprises with sufficient engineering resources, building proprietary coding agents can reduce per-seat costs and provide customization tailored to internal workflows and security requirements. This creates a competitive advantage for larger firms while raising questions about the long-term pricing power of commercial coding AI vendors.

Key Implications

  • Commercial coding AI vendors may face margin pressure and customer concentration risk as larger enterprises build alternatives
  • Access to coding agents is becoming a competitive tool for recruiting and retaining engineering talent
  • Internal tool development requires significant upfront investment, creating a capability gap between well-resourced and smaller firms

What to Watch

Monitor whether this trend spreads to mid-market firms or remains concentrated among large tech companies with dedicated AI teams. Track pricing and feature announcements from Anthropic and OpenAI to see if they respond to internal competition. Watch for any public disclosures of cost savings or performance metrics that could validate the business case for internal development.

Share
AI AgentsAI for BusinessCoding / Dev Tools

Subscribe to the newsletter

The latest stories and analysis, delivered to your inbox.

Free. No spam. Unsubscribe any time.

Related stories

DeepSeek Challenges Claude Code with Open Agent Framework
AI AgentsTrendingModel Release

DeepSeek Challenges Claude Code with Open Agent Framework

DeepSeek launched DeepSeek-V4-Pro, an updated flagship model for agentic workloads, alongside DeepSeek Harness v0.1, an open-source agent framework available under MIT license. The releases position DeepSeek as a competitor to Anthropic's Claude Code and OpenAI's Codex by offering developers an alternative agent infrastructure layer. Simultaneously, DeepSeek is shifting from flat API pricing to peak and off-peak rates starting August 16, with substantially higher prices across the board.

by carl.franzen@venturebeat.com (Carl Franzen)· VentureBeat AI
Google Cuts Gemini Flash Pricing 50% With Faster Iteration
AI AgentsTrendingModel Release

Google Cuts Gemini Flash Pricing 50% With Faster Iteration

Google DeepMind released Gemini 3.7 Flash on August 13, 2026, positioning it as an improved workhorse model for coding and agent-based tasks. The model arrives three weeks after Gemini 3.6 Flash and delivers measurable gains in software engineering, web development, and knowledge-intensive workflows at half the per-token cost of its predecessor. The release reflects developer feedback and algorithmic improvements aimed at production-ready code generation and complex document processing.

· Google Deepmind
Startup Slack Threads Become Commodity for AI Training
AI AgentsTrendingNews

Startup Slack Threads Become Commodity for AI Training

AI training companies like Mercor are actively acquiring internal communications and code from startups, offering payments up to $300,000 for Slack threads, GitHub records, and meeting transcripts. Warmly's CEO received four such acquisition offers within days of the company's HubSpot acquisition announcement. The practice highlights how internal startup data has become a commodity for AI model training, even as acquirers may not want the same datasets.

by Alix Coutures· The Information
Capital One Builds Multi-Agent AI on Customized Open Models
AI AgentsTrendingNews

Capital One Builds Multi-Agent AI on Customized Open Models

Capital One built a multi-agent AI platform centered on customized open-weight models rather than relying on off-the-shelf frontier models. The bank fine-tunes open models with proprietary data and uses a specialized multi-agent orchestration system called MACAW to handle complex workflows like fraud detection and customer service. This approach leverages Capital One's data advantage while enabling extensibility across the enterprise.

· VentureBeat AI