The ORCFLO Indexis an independent benchmark that evaluates large language models the way business professionals actually use them — across real-world tasks spanning analysis, writing, extraction, summarization, and behavioral reliability. Each model is scored on three dimensions (quality, cost, and speed) by a panel of four independent judges. This report evaluates Claude Sonnet 4 in the context of 32 models from ORCFLO Index — May 10, 2026 cohort · 32 total models tested · Anthropic, Google, OpenAI, Mistral.
Claude Sonnet 4
Claude Sonnet 4 lands at #18 of 32 overall with a quality score of 85.53, placing it firmly in the mid-tier of this cohort and well behind frontier leaders like GPT 5 (95.04) and Gemini 3 Pro (93.17). The model shows a sharply uneven profile: elite-tier extraction and near-perfect output consistency are offset by weak summarization and instruction following. A specialist-grade tool whose value depends heavily on workload fit.
Key Findings
- Extraction is the standout capability at 92.0 (rank #4 of 32, leader tier) — competitive with the top of the field despite the model's mid-tier overall placement.
- Output consistency hits 99.0 (rank #2 of 32), making Sonnet 4 one of the most predictable models in the cohort for repeatable workflows.
- Summarization and instruction following lag badly at 73.2 (#29) and 82.0 (#28) respectively, dragging the overall score below several cheaper Anthropic and OpenAI alternatives.
- Cost and speed are middling: $0.0079 per case (#17) and 8.7s response time (#14) — neither a value pick nor a performance leader.
Model Performance: Quality & Cost
The chart below plots quality against cost for all 32 models in the ORCFLO Index. Each dot represents the average quality score a model achieved across the full basket of real-world business tasks, alongside the cost in credits to complete the entire test suite. Models in the upper-left quadrant deliver the highest quality at the lowest cost.
Model Performance: Quality & Time Elapsed
Quality alone doesn’t tell the full story — response time determines whether a model is viable for time-sensitive workflows. The chart below plots each model’s quality score against the total time required to complete the test suite. Models in the upper-left deliver the best quality with the least delay.
Category Scorecard
The ORCFLO Indexevaluates models using real-world business tasks — not academic puzzles or synthetic benchmarks. Each test case is designed to expose specific differences in how models handle the work professionals actually do. Scores are averaged across each category and ranked independently across all 32 models.
| Category | Score | Rank | Tier |
|---|---|---|---|
| Abilities — Core language tasks: what the model can produce when given a well-formed prompt. | |||
| AnalysisReasoning, strategic judgment, disqualifying-factor detection | #21 | Contender | |
| ExtractionField accuracy, null handling, format compliance, zero fabrication | #4 | Leader | |
| SummarizationCompression quality, key-point retention, length compliance | #29 | Trailing | |
| WritingTone, structure, persuasion, audience adaptation | #17 | Contender | |
| Behaviors — How the model acts under pressure: reliability, compliance, and restraint. | |||
| HallucinationFabrication detection, factual grounding, source fidelity | #20 | Contender | |
| Instruction FollowingConstraint adherence, format compliance, multi-part directives | #28 | Trailing | |
| Refusal CalibrationAppropriate refusal vs. over-refusal on legitimate requests | #15 | Strong | |
| Stability — Repeatability and predictability across identical inputs. | |||
| Output ConsistencyRun-to-run reproducibility, format stability, score variance | #2 | Leader | |
Strengths and Cautions
Strengths
- Extraction leader-tier performance at 92.0 (rank #4 of 32) — suitable for structured data pulls, form parsing, and entity-heavy document workflows.
- Near-perfect output consistency at 99.0 (rank #2 of 32), supporting deterministic pipelines where repeatability matters more than raw reasoning.
- Solid refusal calibration at 83.5 (rank #15 of 32, strong tier), indicating balanced judgment on sensitive prompts without excessive over-blocking.
Cautions
- Summarization at 73.2 (rank #29 of 32) sits in trailing tier — avoid for executive briefing generation or document condensation tasks.
- Instruction following at 82.0 (rank #28 of 32) is a material weakness for multi-step or constraint-heavy prompts; expect drift from complex specifications.
- No cost advantage to justify the gaps: at $0.0079 (#17), it is outpriced by Claude Haiku 4.5 ($0.0031) and GPT 4.1 ($0.0052), both of which score within a point overall.
Head-to-Head: Frontier Models
Claude Sonnet 4 is Anthropic’s mid-tier contender in the ORCFLO Index. The table below compares it against the top-performing models from each major provider. Tier assignments use 25% quartiles across the full 32-model field.
| Model | Quality Avg | Quality Rank | Cost Rank | Speed Rank |
|---|---|---|---|---|
| o3 | 86.2 | #15 | #26 | #22 |
| GPT 5 Nano | 85.9 | #16 | #10 | #30 |
| o4-mini | 85.9 | #17 | #21 | #12 |
| Claude Sonnet 4 | 85.5 | #18 | #17 | #14 |
| GPT 4.1 | 85.0 | #19 | #15 | #10 |
| Claude Haiku 4.5 | 84.7 | #20 | #13 | #11 |
| Claude Sonnet 4.5 | 84.6 | #21 | #20 | #17 |
When to Use Claude Sonnet 4
The ORCFLO Index
This evaluation covers 40 cases across 8 categories. All tasks are text-only and English-only. Code generation, multi-turn conversation, multimodal tasks, and agentic workflows are not tested. Each contestant is scored by a panel of four independent judges — Gemini 2.5 Pro, Claude Opus 4.7, GPT 5.5, and Mistral Large — with final scores averaged across all four. Cost and speed measurements reflect API pricing and latency as of the test date (May 10, 2026) and will change as providers update their offerings.
How We Test
The ORCFLO Indexevaluates large language models across three independent dimensions — quality, cost, and speed — using real-world business tasks designed to expose the differences that matter for model selection. Each model is scored by a panel of four independent judges to reduce single-model bias.
- Test Cases
- 40 cases across 8 categories spanning Abilities (Analysis, Extraction, Summarization, Writing), Behaviors (Hallucination, Instruction Following, Refusal Calibration), and Stability (Output Consistency).
- Judge Panel
- Gemini 2.5 Pro, Claude Opus 4.7, GPT 5.5, and Mistral Large. Each judge scores independently. Final score is the average across all four.
- Scoring
- Three independent ranks: quality (higher is better), cost (lower is better), speed (faster is better). No composite score — composites hide the tradeoffs that drive model-selection decisions.
- Tier Definitions
- LeaderQuality ≥ 90.8Ranks 1–8Strong≥ 85.9Ranks 9–16Contender≥ 80.9Ranks 17–24Trailing< 80.9Ranks 25–32