GPT-4o vs Claude Haiku 3.5 vs Llama-3.2-90B-Vision-Instruct
Too close to call on our weighted score (GPT-4o 51, Claude Haiku 3.5 48, Llama-3.2-90B-Vision-Instruct 44). The right pick depends on what you value most.
OpenAIGPT-4o
51/100- ECI129.0
- Price$2.50 / $10.00
- Context128K
AnthropicClaude Haiku 3.5
48/100- ECI127.0
- Price$0.80 / $4.00
- Context200K
MetaLlama-3.2-90B-Vision-Instruct
44/100- ECI125.5
- Price—
- Context128K
Too close to call
It is close. Our weighted score puts them within 3 points (GPT-4o 51/100, Claude Haiku 3.5 48/100, Llama-3.2-90B-Vision-Instruct 44/100), so choose by what matters most for your work: GPT-4o for raw capability, Claude Haiku 3.5 on price and Claude Haiku 3.5 for long inputs. The score weighs capability 67%, inputs & features 20%, context window 13%.
- CapabilityGPT-4oCapabilities Index (ECI): GPT-4o 129.0 · Claude Haiku 3.5 127.0 · Llama-3.2-90B-Vision-Instruct 125.5
- Lowest priceClaude Haiku 3.5Claude Haiku 3.5 $1.60 · GPT-4o $4.38 per 1M tokens (3:1 blend) · Llama-3.2-90B-Vision-Instruct unpriced
- Longest contextClaude Haiku 3.5Claude Haiku 3.5 200,000 · GPT-4o 128,000 · Llama-3.2-90B-Vision-Instruct 128,000 tokens
- Widest inputsGPT-4o and Claude Haiku 3.5GPT-4o: Text, Images, PDFs · Claude Haiku 3.5: Text, Images, PDFs · Llama-3.2-90B-Vision-Instruct: Text, Images
- Self-hostingLlama-3.2-90B-Vision-InstructPublishes downloadable weights (Llama 3.2 Community License)
| Measure | Weight | GPT-4o | Claude Haiku 3.5 | Llama-3.2-90B-Vision-Instruct |
|---|---|---|---|---|
| CapabilityCapabilities Index (ECI) | 67% | 51 | 48 | 46 |
| Inputs & features | 20% | 70 | 60 | 50 |
| Context window | 13% | 24 | 32 | 24 |
| Overall | 100% | 51/100 | 48/100 | 44/100 |
Left out because at least one model lacks the data: price. The remaining weights were rescaled.
Every spec in one table
Highlighted cells lead their row. Dashes mean the data is not published.
| Specification | GPT-4oOpenAI | Claude Haiku 3.5Anthropic | Llama-3.2-90B-Vision-InstructMeta |
|---|---|---|---|
| Capability | |||
| Capabilities Index (ECI) | 129.0 (best) | 127.0 | 125.5 |
| ECI rank | #141 of 165 (best) | #147 of 165 | #150 of 165 |
| GPQA DiamondGraduate-level science questions | 48.9% (best) | 38.1% | 41.0% |
| OTIS Mock AIME 2024–2025Competition mathematics | 6.3% (best) | 4.3% | 2.6% |
| Price per million tokens | |||
| Input | $2.50 | $0.80 (best) | — |
| Output | $10.00 | $4.00 (best) | — |
| Cached input | $1.25 | — | — |
| Blended (3:1) | $4.38 | $1.60 (best) | — |
| Long-context rate | Same rate | Same rate | — |
| Price source | Official OpenAI API | Median of 3 providers | — |
| Limits | |||
| Context window | 128,000 tokens | 200,000 tokens (best) | 128,000 tokens |
| Max output | 16,384 tokens (best) | 8,192 tokens | 4,096 tokens |
| Inputs and features | |||
| Text | Yes | Yes | Yes |
| Images | Yes | Yes | Yes |
| PDFs | Yes | Yes | No |
| Audio | No | No | No |
| Video | No | No | No |
| Reasoning | No | No | No |
| Tool calling | Yes | Yes | Yes |
| Structured output | Yes | No | No |
| Availability | |||
| Weights | Proprietary | Proprietary | OpenLlama 3.2 Community License |
| API model ID | gpt-4o | — | — |
| API providers | 26 (best) | 4 | 1 |
| Released | May 13, 2024 | Oct 22, 2024 | Sep 25, 2024 |
| Knowledge cutoff | Sep 2023 | Jul 31, 2024 | Dec 2023 |
What would a month cost?
Enter your expected volume in millions of tokens. List prices only; caching and batch discounts would lower these.
- GPT-4o$45.00
- Claude Haiku 3.5$16.00
- Llama-3.2-90B-Vision-Instruct—
Which should you choose?
Which is better: GPT-4o, Claude Haiku 3.5 or Llama-3.2-90B-Vision-Instruct?
It is close. Our weighted score puts them within 3 points (GPT-4o 51/100, Claude Haiku 3.5 48/100, Llama-3.2-90B-Vision-Instruct 44/100), so choose by what matters most for your work: GPT-4o for raw capability, Claude Haiku 3.5 on price and Claude Haiku 3.5 for long inputs. The score weighs capability 67%, inputs & features 20%, context window 13%.
Which is cheaper, GPT-4o, Claude Haiku 3.5 or Llama-3.2-90B-Vision-Instruct?
Claude Haiku 3.5 is cheaper at $0.80 input / $4.00 output per million tokens (median across 3 API providers). GPT-4o costs $2.50 input / $10.00 output per million tokens (official OpenAI API price). At a typical mix of three input tokens to one output token, that is $1.60 per million tokens for Claude Haiku 3.5 versus $4.38 for GPT-4o (2.7× as much). Llama-3.2-90B-Vision-Instruct has no published per-token price.
Which scores higher on benchmarks?
GPT-4o scores higher on the Capabilities Index (ECI): GPT-4o 129.0 (#141 of 165), Claude Haiku 3.5 127.0 (#147 of 165) and Llama-3.2-90B-Vision-Instruct 125.5 (#150 of 165). The confidence ranges of the top two overlap (124.7–131.3 vs 121.4–129.2), so treat the gap as small. On individual benchmarks: GPQA Diamond — GPT-4o 48.9%, Llama-3.2-90B-Vision-Instruct 41.0%, Claude Haiku 3.5 38.1%; OTIS Mock AIME 2024–2025 — GPT-4o 6.3%, Claude Haiku 3.5 4.3%, Llama-3.2-90B-Vision-Instruct 2.6%.
Which is better for coding?
There are no published SWE-bench Verified results for GPT-4o, Claude Haiku 3.5 and Llama-3.2-90B-Vision-Instruct yet, so there is no like-for-like coding score. On overall capability, GPT-4o leads, which tends to carry over to coding, but test on your own codebase. All three support tool calling for agent workflows.
Which has the bigger context window?
Claude Haiku 3.5 has the largest context window at 200,000 tokens, against 128,000 for GPT-4o and 128,000 for Llama-3.2-90B-Vision-Instruct. Maximum output per response: GPT-4o up to 16,384, Claude Haiku 3.5 up to 8,192, Llama-3.2-90B-Vision-Instruct up to 4,096 tokens.
Which can read images, PDFs, audio or video?
GPT-4o accepts text, images and PDFs; Claude Haiku 3.5 accepts text, images and PDFs; Llama-3.2-90B-Vision-Instruct accepts text and images. GPT-4o handles the widest range of inputs.
Are any of these open source?
Llama-3.2-90B-Vision-Instruct publishes its weights (Llama 3.2 Community License) and can be self-hosted; GPT-4o and Claude Haiku 3.5 is proprietary.
Which is newer?
Claude Haiku 3.5 is the newest, released Oct 22, 2024. Llama-3.2-90B-Vision-Instruct came out Sep 25, 2024; GPT-4o came out May 13, 2024. Knowledge cutoff: GPT-4o Sep 2023, Claude Haiku 3.5 Jul 31, 2024, Llama-3.2-90B-Vision-Instruct Dec 2023.
How do you decide the winner?
Each model gets a 0–100 score on capability (50%, independent benchmark results); price (25%, blended price per million tokens (3 input : 1 output), log scale); inputs & features (15%, image, PDF, audio and video input, tool calling, structured output and reasoning); context window (10%, maximum tokens per request, log scale). Dimensions missing for any model are dropped and the remaining weights rescaled, so every model is judged on the same evidence. Specs and prices come from public model listings and the labs’ own API pages; capability scores come from independent benchmark runs. Data updated Nov 2, 2026.
