AnthropicEndsClaudeSubscriptionSubsidy,EvalCostsSpike

◆ DEEP DIVES

1. 01

   Anthropic's Pricing Earthquake: Three Simultaneous Cost Shocks Hit Your Claude Stack

   What Happened

   The headline change: Claude subscriptions now convert to dollar-matched API credits for all programmatic usage, including Agent SDK, claude-p, GitHub Actions, and third-party harnesses. The implicit 70-90% effective discount is gone. In the same week, Opus 4.7 tripled image processing cost, and starting June 15, third-party tool usage (Zed, Conductor, OpenCode, T3 Code) moves to a separate credit bucket equal to plan value with no rollover and overflow at API rates.

   The driver behind the pricing is capacity. Anthropic planned for 10x growth and is seeing 80x. The emergency fix is leasing xAI's entire Colossus 1 cluster, 220,000+ GPUs spanning H100, H200, and GB200. A CFO is in seat and the company is targeting an October IPO, which is a reasonable proxy for why margin-per-token is now a board metric.

   ---

   The Capacity Context

   The pricing changes read more cleanly alongside the capacity numbers. ServiceNow's CDIO burned through a full-year Claude budget by May. National Life Group's CIO called Claude 'not great for companies that want per-user monitoring,' and Anthropic ships no native per-user telemetry and no SLAs, which is unusual for a dependency sitting on production critical paths.

   Change	Impact	Timeline
   Subscription → API credits	70-90% discount gone on batch/eval workloads	Immediate
   Opus 4.7 image cost	3x on multimodal pipelines	Immediate
   June 15 third-party split	No subsidized tokens for Zed/OpenCode/T3	30 days
   Colossus integration	p95/p99 variance during heterogeneous fleet merge	Weeks–months

   Any Claude benchmark from before May 7 is stale, and any cost model built against flat subscription rates is not directionally wrong, it is numerically wrong.

   ---

   The OpenAI Counter-Move

   Sam Altman posted a 2-month-free Codex enterprise switch promo on the same day Anthropic metered subscriptions. Ramp's April data showed Anthropic edging OpenAI for the first time, 34.4% vs 32.3%. The promo is an asymmetric-payoff bet: free to evaluate, with bounded switching cost if you already have a provider abstraction layer. OpenAI is pricing directly into the developer cohort Anthropic just alienated.

   Cross-Source Tension

   Sources disagree on the durability of Anthropic's lead. Ramp data is a card-spend proxy and measures who gets billed, not token volume or production criticality. A 20-seat pilot weighs the same as a company at inference scale, and OpenAI notes large enterprises rarely pay by card. The directional signal looks real; the magnitude does not. What is not uncertain is that the market is now genuinely multi-vendor, and architecture should reflect that.

   Action items

   • Audit every Claude-backed workload (Agent SDK, GitHub Actions, batch evals) and reconcile projected token burn against the new credit cap by end of this week
   • Deploy an LLM gateway (LiteLLM, Portkey) with per-user, per-feature tagging and daily budget alerts within this sprint
   • Run a 2-month Codex evaluation under OpenAI's enterprise switch promo with matched prompts against your Claude harness
   • Reforecast Claude inference spend for any team using Zed/Conductor/OpenCode modeling the post-June-15 scenario

   Sources:Claude just metered your agent SDK calls · Claude Code latency on long-context requests drifted upward · Anthropic ships no per-user usage telemetry · Anthropic passes OpenAI in B2B · Vercel published a number worth sitting with · Ramp's AI Index shows Anthropic at 34.4%

2. 02

   59% Agentic Volume: Your Eval Harness and Cost Model Are Measuring the Minority

   The Production Shift, Quantified

   Vercel's AI Gateway telemetry — 200,000 teams over 7 months — reports that 59% of all token volume is now agentic: multi-turn, tool-calling traces, not single-shot completions. Six months ago the share was under 20%. The interesting structure is the spend-volume split. Anthropic captures 61% of spend via Opus on reasoning and planning nodes, while Google captures 38% of volume via Flash on throughput and utility calls. Vendor loyalty is not visible in the data. Customers churn freely.

   This is not a leaderboard result. It is production telemetry from a multi-tenant gateway, which means most tokens in the wild are inside multi-step tool loops with retries, not the single-turn completions an eval harness was built to score.

   ---

   What Breaks at 59%

   Eval harnesses: single-turn accuracy on held-out prompts measures the 41% minority. The thing this doesn't tell you is whether a planner burns 40K tokens arguing with itself before giving up. Pass rate looks fine. Spend is 5x budget. Cost models: most were calibrated when input-output ratios sat at 3:1. Agentic traces run at ~15:1 on input, with heavy cache reuse on some providers and none on others. A forecast carried over from last year's ratio is off by roughly 5x on spend.

   Metric	Single-Turn World	59% Agentic World
   Input:Output ratio	~3:1	~15:1
   Cost driver	Output tokens	Trajectory length × retries
   Eval metric	Accuracy on final answer	Cost-per-successful-task
   Routing unit	Single request	Session with KV cache state
   Failure mode	Wrong answer	Correct answer at 10x budget

   If 59% of your tokens are agentic but 100% of your evals are single-turn, you're flying instruments-out — update the harness before you update the model.

   ---

   The Routing Architecture That Fits

   The Vercel pattern is consistent with what Abridge disclosed across 80M+ clinical conversations: a constellation of models, cheap fast triage in front, expensive reasoning behind, per-task selection. Given those numbers, the plausible envelope for cost reduction from tiered routing is 20-40% at constant trajectory completion rate. Glean's published benchmark puts off-the-shelf MCP at 30% more tokens than retrieval-tuned knowledge graphs. The methodology is not disclosed and the source is the vendor, so treat it as directional. It is consistent with the MCP context-window bloat people have measured independently.

   The Enterprise Convergence

   SAP (€100M partner investment) and ServiceNow (Action Fabric) have landed in the same place: agents need Knowledge Graph grounding + MCP-exposed workflows. The architectural drift is from RAG-over-docs to structured, entity-resolved context. The eval needs to move with it. Tool-use accuracy against grounded context is the production differentiator. It is on no public leaderboard.

   Action items

   • Add trajectory-level metrics to your eval harness this sprint: tool-call F1, steps-to-completion, cost-per-successful-task, recovery-from-error rate
   • Instrument per-node token cost across your agent graphs and route utility calls (summarization, extraction, query rewriting) to Flash/Haiku-class models within 2 weeks
   • Run a 1-hour spike measuring token overhead of current MCP/tool-calling setup vs. a retrieval-first baseline on 100 production traces
   • Add LLM-judge ↔ human-annotator agreement as a tracked SLI this quarter; re-calibrate when judge model changes

   Sources:Agentic traffic crossed fifty-nine percent · Vercel published a number worth sitting with · AI Gateway data puts agentic workloads at fifty-nine percent · Abridge runs model routing across 100M conversations · MCP plus knowledge graphs

3. 03

   Training Efficiency Frontier: Three Results That Change Unit Economics This Quarter

   Three Independent Wins, One Direction

   The marginal dollar in model development has moved from raw compute to architecture and curation. The evidence this week comes from three separate labs, each working a different layer of the training stack, and the results point the same way. The marginal dollar in model development has moved from raw compute to architecture and curation.

   Work	Claim	Scale Validated	Inference Impact	Replication Risk
   Nous Research TST	2-3x wall-clock at matched FLOPs	270M → 10B-A1B MoE	None — no architecture change	Medium; single-source, clean claim
   NVIDIA Star Elastic	360x cheaper model-family derivation; 7x vs SOTA compression	Not specified	Produces family of sizes from one run	High; big number, lab-reported
   Datology VLM	+11.7 pts on 20 benchmarks at 17x less compute	2B and 4B params	3.3x lower response FLOPs (serving win)	Medium; benchmark selection risk

   ---

   TST Is the One to Spike First

   Token Superposition Training is a pretraining recipe change with no inference-side downstream. If it replicates, it is a 2-3x improvement on wall-clock for free. The serving architecture is unchanged. The efficiency lives entirely in how gradients are computed during training. Validated from 270M up to a 10B-A1B MoE, which is the band most teams actually run for continued pretraining and domain adaptation.

   Datology: Curation Beats Compute

   Datology's result is the cleanest evidence so far that data curation now dominates compute scaling for VLMs. A 2B model lands about 10 points above InternVL3.5-2B at 17x less training compute, on data selection alone. At 4B, they reach near-frontier quality at 3.3x lower response FLOPs than Qwen3-VL-4B, which is a serving win and not just a training one. For any team maintaining a vision pipeline, this reorders the next budget line. Curation tooling before more GPUs.

   Star Elastic: Speculative but High-Leverage

   NVIDIA's claim that one post-training run produces a full family of model sizes at 360x lower cost than pretraining each is the kind of number that always shrinks under independent evaluation. The thing this number doesn't tell you is what holds at the small end versus the large end. Even a 30x hold would restructure how teams produce size tiers for routing. Paired with the 59% agentic routing pattern, cheap access to a 1B-to-70B spectrum from a single run is what a tiered architecture actually needs.

   TST is worth a spike on your next continued-pretraining run. If wall-clock comes in at even 1.6x with no val-loss regression, it pays for itself immediately.

   ---

   The Parallel Signal: Only 15% Are Ready

   Fivetran's readiness index reports that 15% of organizations have the data foundation for agentic AI. Data quality and lineage are the top blocker, cited by roughly 50%. Read against the efficiency results, the picture is consistent. Training is getting cheaper. Most teams cannot capitalize because the data layer is not there. Half the agent projects funded this quarter are data-platform projects with an agent bolted on.

   Action items

   • Spike Token Superposition Training on a 1B-param continued-pretraining run against a matched-FLOPs baseline within the next 2 weeks
   • Run an ablation benchmarking Qwen-2.5/3 or DeepSeek-V3 against your current production model on in-domain evals this quarter
   • Audit ANALYZE/compute-stats coverage across top-20 Iceberg/Delta tables; add stats freshness to table-level SLAs
   • Score target domains against Fivetran readiness dimensions (quality, lineage, governance) before greenlighting agentic AI projects

   Sources:Claude just metered your agent SDK calls · DuckDB shipped a client-server mode this week