GPT-5
Proprietary CommercialDeveloped by OpenAI · Unified Omni-Reasoning MoE · Modalities: Text, Vision, Audio, Video · Released 2025-11-18
What are the token costs and operational benchmarks for GPT-5?
Verified Capability Benchmarks
Composite Intelligence: 85.6/100Architectural & Production Profile
OpenAI's unified flagship system combining deep test-time reasoning with native multimodal perception and 400k context.
Complex enterprise automation, full-stack software engineering, multimodal agent orchestration, and scientific synthesis.
- Architecture
- Unified Omni-Reasoning MoE
- Modalities
- Text, Vision, Audio, Video
- Max Output Tokens
- 128,000 tokens
- Blended Rate (3:1)
- $4.375 / 1M
Verified Cloud Inference Endpoints Hosting GPT-5
Compare latency, streaming speed, and token pricing across cloud API providers.
| Inference Provider | Role | Input / 1M | Output / 1M | Avg TTFT | Throughput |
|---|---|---|---|---|---|
| OpenAI Platform API → | Primary Host | $2.50 | $10.00 | 340ms | 105 tok/s |
| Anysphere Cursor Router → | Partner Endpoint | $2.50 | $10.00 | 360ms | 105 tok/s |
Interactive Monthly Token Economics & ROI Forecaster
Model production workloads across prompt, completion, and cached context tokens, or switch models on the fly.
Estimated Monthly Spend
$2.50/1M in · $10.00/1M out
Prompt Cache: $0.625/1M read · $2.50/1M write
$2.00/1M in · $10.00/1M out
Prompt Cache: $0.100/1M read · $0.50/1M write
Head-to-Head Comparisons Involving GPT-5 (36)
GPT-5 vs Gemini 4 Argon
Gemini 4 Argon delivers superior capability and a 20.0% input cost advantage over GPT-5. At $2.00 vs. $2.50 per 1M input tokens ($10.00 output), Gemini leads SWE-bench Verified (84.5% vs. 74.9%) and GPQA Diamond (93.8% vs. 85.4%). Conversely, GPT-5 posts a 340ms TTFT, responding 510ms faster despite lower throughput (105 vs. 115 tok/s).
GPT-5 vs Claude 4.5 Sonnet
GPT-5 is 16.7% cheaper for input tokens ($2.50 vs. $3.00 per 1M tokens) and $10.00 vs. $15.00 for output tokens (1.2x cost advantage). In terms of operational performance, GPT-5 delivers lower response latency at 340ms TTFT (40ms faster than Claude 4.5 Sonnet) with 105 vs. 84 tok/s throughput. Claude 4.5 Sonnet leads coding evaluations at 75.8% SWE-bench Verified vs. 74.9% for GPT-5 and 84.2% vs. 85.4% on GPQA Diamond.
GPT-5 vs Claude Opus 5.5
GPT-5 is 37.5% cheaper for input tokens ($2.50 vs. $4.00 per 1M tokens) and $10.00 vs. $20.00 for output tokens (1.6x cost advantage). In terms of operational performance, GPT-5 delivers lower response latency at 340ms TTFT (340ms faster than Claude Opus 5.5) with 105 vs. 95 tok/s throughput. Claude Opus 5.5 leads coding evaluations at 89.9% SWE-bench Verified vs. 74.9% for GPT-5 and 93.2% vs. 85.4% on GPQA Diamond.
GPT-5 vs Gemini 3.8 Flash
Gemini 3.8 Flash is 90.0% cheaper for input tokens ($0.25 vs. $2.50 per 1M tokens) and $1.50 vs. $10.00 for output tokens (10.0x cost advantage). In terms of operational performance, Gemini 3.8 Flash delivers lower response latency at 105ms TTFT (235ms faster than GPT-5) with 210 vs. 105 tok/s throughput. GPT-5 leads coding evaluations at 74.9% SWE-bench Verified vs. 68.5% for Gemini 3.8 Flash and 85.4% vs. 79.2% on GPQA Diamond.
GPT-5 vs GPT-6 Astra
GPT-5 is 75.0% cheaper for input tokens ($2.50 vs. $10.00 per 1M tokens) and $10.00 vs. $50.00 for output tokens (4.0x cost advantage). In terms of operational performance, GPT-5 delivers lower response latency at 340ms TTFT (400ms faster than GPT-6 Astra) with 105 vs. 74 tok/s throughput. GPT-6 Astra leads coding evaluations at 75.7% SWE-bench Verified vs. 74.9% for GPT-5 and 96.0% vs. 85.4% on GPQA Diamond.
GPT-5 vs DeepSeek V4 Pro
DeepSeek V4 Pro is 47.2% cheaper for input tokens ($1.32 vs. $2.50 per 1M tokens) and $3.96 vs. $10.00 for output tokens (1.9x cost advantage). In terms of operational performance, GPT-5 delivers lower response latency at 340ms TTFT (180ms faster than DeepSeek V4 Pro) with 105 vs. 78 tok/s throughput. DeepSeek V4 Pro leads coding evaluations at 80.6% SWE-bench Verified vs. 74.9% for GPT-5 and 90.1% vs. 85.4% on GPQA Diamond.
GPT-5 vs DeepSeek R1
DeepSeek R1 is 78.0% cheaper for input tokens ($0.55 vs. $2.50 per 1M tokens) and $2.19 vs. $10.00 for output tokens (4.5x cost advantage). In terms of operational performance, GPT-5 delivers lower response latency at 340ms TTFT (1110ms faster than DeepSeek R1) with 105 vs. 58 tok/s throughput. GPT-5 leads coding evaluations at 74.9% SWE-bench Verified vs. 49.2% for DeepSeek R1 and 85.4% vs. 71.5% on GPQA Diamond.
GPT-5 vs DeepSeek V4 Flash Vision (Exp)
DeepSeek V4 Flash Vision (Exp) outperforms GPT-5 with an 11.4x cost advantage and superior capability. It cuts input pricing by 91.2% ($0.22 vs. $2.50) and output to $0.65 vs. $10.00 per 1M tokens. DeepSeek also achieves a faster 310ms TTFT and 165 tok/s, while leading SWE-bench Verified (77.4% vs. 74.9%) and GPQA Diamond (91.3% vs. 85.4%).
GPT-5 vs OpenAI o3
OpenAI o3 is 20.0% cheaper for input tokens ($2.00 vs. $2.50 per 1M tokens) and $8.00 vs. $10.00 for output tokens (1.2x cost advantage). In terms of operational performance, GPT-5 delivers lower response latency at 340ms TTFT (810ms faster than OpenAI o3) with 105 vs. 82 tok/s throughput. GPT-5 leads coding evaluations at 74.9% SWE-bench Verified vs. 71.7% for OpenAI o3 and 85.4% vs. 87.7% on GPQA Diamond.
GPT-5 vs Claude 4 Opus
GPT-5 is 83.3% cheaper for input tokens ($2.50 vs. $15.00 per 1M tokens) and $10.00 vs. $75.00 for output tokens (6.0x cost advantage). In terms of operational performance, GPT-5 delivers lower response latency at 340ms TTFT (580ms faster than Claude 4 Opus) with 105 vs. 46 tok/s throughput. Claude 4 Opus leads coding evaluations at 77.2% SWE-bench Verified vs. 74.9% for GPT-5 and 87.1% vs. 85.4% on GPQA Diamond.
GPT-5 vs Claude 3.7 Sonnet
GPT-5 is 16.7% cheaper for input tokens ($2.50 vs. $3.00 per 1M tokens) and $10.00 vs. $15.00 for output tokens (1.2x cost advantage). In terms of operational performance, GPT-5 delivers lower response latency at 340ms TTFT (200ms faster than Claude 3.7 Sonnet) with 105 vs. 72 tok/s throughput. GPT-5 leads coding evaluations at 74.9% SWE-bench Verified vs. 70.3% for Claude 3.7 Sonnet and 85.4% vs. 78.5% on GPQA Diamond.
GPT-5 vs Gemini 3.1 Pro
Gemini 3.1 Pro is 50.0% cheaper for input tokens ($1.25 vs. $2.50 per 1M tokens) and $10.00 vs. $10.00 for output tokens (2.0x cost advantage). In terms of operational performance, GPT-5 delivers lower response latency at 340ms TTFT (80ms faster than Gemini 3.1 Pro) with 105 vs. 98 tok/s throughput. GPT-5 leads coding evaluations at 74.9% SWE-bench Verified vs. 71.8% for Gemini 3.1 Pro and 85.4% vs. 82.4% on GPQA Diamond.
GPT-5 vs Grok 4
GPT-5 and Grok 4 share identical input pricing at $2.50 per 1M tokens ($10.00 vs. $12.50 output). In terms of operational performance, GPT-5 delivers lower response latency at 340ms TTFT (140ms faster than Grok 4) with 105 vs. 95 tok/s throughput. GPT-5 leads coding evaluations at 74.9% SWE-bench Verified vs. 73.2% for Grok 4 and 85.4% vs. 86.2% on GPQA Diamond.
GPT-5 vs Llama 4 Maverick
Llama 4 Maverick is 92.0% cheaper for input tokens ($0.20 vs. $2.50 per 1M tokens) and $0.60 vs. $10.00 for output tokens (12.5x cost advantage). In terms of operational performance, Llama 4 Maverick delivers lower response latency at 135ms TTFT (205ms faster than GPT-5) with 175 vs. 105 tok/s throughput. GPT-5 leads coding evaluations at 74.9% SWE-bench Verified vs. 62.4% for Llama 4 Maverick and 85.4% vs. 73.8% on GPQA Diamond.
GPT-5 vs Qwen 3.8 Flash Next
Qwen 3.8 Flash Next outperforms GPT-5 in capability while delivering a 15.6x cost advantage. Input costs 93.6% less ($0.16 vs. $2.50 per 1M) with $0.47 vs. $10.00 output pricing. Qwen also achieves lower latency (280ms TTFT, 60ms faster), higher throughput (185 vs. 105 tok/s), and leads SWE-bench Verified (81.0% vs. 74.9%) and GPQA Diamond (91.7% vs. 85.4%).
GPT-5 vs Gemini 3.5 Flash-Lite
Gemini 3.5 Flash-Lite is 97.0% cheaper for input tokens ($0.07 vs. $2.50 per 1M tokens) and $0.30 vs. $10.00 for output tokens (33.3x cost advantage). In terms of operational performance, Gemini 3.5 Flash-Lite delivers lower response latency at 78ms TTFT (262ms faster than GPT-5) with 265 vs. 105 tok/s throughput. GPT-5 leads coding evaluations at 74.9% SWE-bench Verified vs. 55.4% for Gemini 3.5 Flash-Lite and 85.4% vs. 68.4% on GPQA Diamond.
GPT-5 vs OpenAI o4-mini
OpenAI o4-mini is 56.0% cheaper for input tokens ($1.10 vs. $2.50 per 1M tokens) and $4.40 vs. $10.00 for output tokens (2.3x cost advantage). In terms of operational performance, GPT-5 delivers lower response latency at 340ms TTFT (340ms faster than OpenAI o4-mini) with 105 vs. 125 tok/s throughput. GPT-5 leads coding evaluations at 74.9% SWE-bench Verified vs. 68.1% for OpenAI o4-mini and 85.4% vs. 81.4% on GPQA Diamond.
GPT-5 vs DeepSeek V3
DeepSeek V3 is 94.4% cheaper for input tokens ($0.14 vs. $2.50 per 1M tokens) and $0.28 vs. $10.00 for output tokens (17.9x cost advantage). In terms of operational performance, GPT-5 delivers lower response latency at 340ms TTFT (70ms faster than DeepSeek V3) with 105 vs. 78 tok/s throughput. GPT-5 leads coding evaluations at 74.9% SWE-bench Verified vs. 42.0% for DeepSeek V3 and 85.4% vs. 59.1% on GPQA Diamond.
GPT-5 vs Gemma 4 31B IT
Gemma 4 31B IT is 95.2% cheaper for input tokens ($0.12 vs. $2.50 per 1M tokens) and $0.36 vs. $10.00 for output tokens (20.8x cost advantage). In terms of operational performance, Gemma 4 31B IT delivers lower response latency at 115ms TTFT (225ms faster than GPT-5) with 155 vs. 105 tok/s throughput. GPT-5 leads coding evaluations at 74.9% SWE-bench Verified vs. 59.2% for Gemma 4 31B IT and 85.4% vs. 71.2% on GPQA Diamond.
GPT-5 vs OpenAI o3-mini
OpenAI o3-mini is 56.0% cheaper for input tokens ($1.10 vs. $2.50 per 1M tokens) and $4.40 vs. $10.00 for output tokens (2.3x cost advantage). In terms of operational performance, GPT-5 delivers lower response latency at 340ms TTFT (610ms faster than OpenAI o3-mini) with 105 vs. 98 tok/s throughput. GPT-5 leads coding evaluations at 74.9% SWE-bench Verified vs. 49.3% for OpenAI o3-mini and 85.4% vs. 79.7% on GPQA Diamond.
GPT-5 vs Claude 3.5 Haiku
Claude 3.5 Haiku is 68.0% cheaper for input tokens ($0.80 vs. $2.50 per 1M tokens) and $4.00 vs. $10.00 for output tokens (3.1x cost advantage). In terms of operational performance, Claude 3.5 Haiku delivers lower response latency at 190ms TTFT (150ms faster than GPT-5) with 130 vs. 105 tok/s throughput. GPT-5 leads coding evaluations at 74.9% SWE-bench Verified vs. 40.6% for Claude 3.5 Haiku and 85.4% vs. 41.6% on GPQA Diamond.
GPT-5 vs Claude 3.5 Sonnet
GPT-5 is 16.7% cheaper for input tokens ($2.50 vs. $3.00 per 1M tokens) and $10.00 vs. $15.00 for output tokens (1.2x cost advantage). In terms of operational performance, Claude 3.5 Sonnet delivers lower response latency at 310ms TTFT (30ms faster than GPT-5) with 80 vs. 105 tok/s throughput. GPT-5 leads coding evaluations at 74.9% SWE-bench Verified vs. 49.0% for Claude 3.5 Sonnet and 85.4% vs. 59.4% on GPQA Diamond.
GPT-5 vs Gemini 2.0 Flash
Gemini 2.0 Flash is 96.0% cheaper for input tokens ($0.10 vs. $2.50 per 1M tokens) and $0.40 vs. $10.00 for output tokens (25.0x cost advantage). In terms of operational performance, Gemini 2.0 Flash delivers lower response latency at 280ms TTFT (60ms faster than GPT-5) with 185 vs. 105 tok/s throughput. GPT-5 leads coding evaluations at 74.9% SWE-bench Verified vs. 51.8% for Gemini 2.0 Flash and 85.4% vs. 63.6% on GPQA Diamond.
GPT-5 vs Gemini 2.5 Flash
Gemini 2.5 Flash is 88.0% cheaper for input tokens ($0.30 vs. $2.50 per 1M tokens) and $2.50 vs. $10.00 for output tokens (8.3x cost advantage). In terms of operational performance, Gemini 2.5 Flash delivers lower response latency at 110ms TTFT (230ms faster than GPT-5) with 180 vs. 105 tok/s throughput. GPT-5 leads coding evaluations at 74.9% SWE-bench Verified vs. 58.0% for Gemini 2.5 Flash and 85.4% vs. 70.5% on GPQA Diamond.
GPT-5 vs Gemini 2.5 Pro
Gemini 2.5 Pro is 50.0% cheaper for input tokens ($1.25 vs. $2.50 per 1M tokens) and $10.00 vs. $10.00 for output tokens (2.0x cost advantage). In terms of operational performance, GPT-5 delivers lower response latency at 340ms TTFT (360ms faster than Gemini 2.5 Pro) with 105 vs. 80 tok/s throughput. GPT-5 leads coding evaluations at 74.9% SWE-bench Verified vs. 63.8% for Gemini 2.5 Pro and 85.4% vs. 74.0% on GPQA Diamond.
GPT-5 vs GLM-5.3-Flash
GLM-5.3-Flash is 95.2% cheaper for input tokens ($0.12 vs. $2.50 per 1M tokens) and $0.40 vs. $10.00 for output tokens (20.8x cost advantage). In terms of operational performance, GLM-5.3-Flash delivers lower response latency at 90ms TTFT (250ms faster than GPT-5) with 175 vs. 105 tok/s throughput. GPT-5 leads coding evaluations at 74.9% SWE-bench Verified vs. 52.8% for GLM-5.3-Flash and 85.4% vs. 63.5% on GPQA Diamond.
GPT-5 vs GLM-5.3
GLM-5.3 is 86.0% cheaper for input tokens ($0.35 vs. $2.50 per 1M tokens) and $1.00 vs. $10.00 for output tokens (7.1x cost advantage). In terms of operational performance, GLM-5.3 delivers lower response latency at 160ms TTFT (180ms faster than GPT-5) with 120 vs. 105 tok/s throughput. GPT-5 leads coding evaluations at 74.9% SWE-bench Verified vs. 56.0% for GLM-5.3 and 85.4% vs. 67.0% on GPQA Diamond.
GPT-5 vs GPT-4o Mini
GPT-4o Mini is 94.0% cheaper for input tokens ($0.15 vs. $2.50 per 1M tokens) and $0.60 vs. $10.00 for output tokens (16.7x cost advantage). In terms of operational performance, GPT-4o Mini delivers lower response latency at 220ms TTFT (120ms faster than GPT-5) with 140 vs. 105 tok/s throughput. GPT-5 leads coding evaluations at 74.9% SWE-bench Verified vs. 41.0% for GPT-4o Mini and 85.4% vs. 40.2% on GPQA Diamond.
GPT-5 vs GPT-4o
GPT-4o and GPT-5 share identical input pricing at $2.50 per 1M tokens ($10.00 vs. $10.00 output). In terms of operational performance, GPT-4o delivers lower response latency at 280ms TTFT (60ms faster than GPT-5) with 100 vs. 105 tok/s throughput. GPT-5 leads coding evaluations at 74.9% SWE-bench Verified vs. 48.0% for GPT-4o and 85.4% vs. 53.6% on GPQA Diamond.
GPT-5 vs GPT-6.1 Sol
GPT-6.1 Sol is 20.0% cheaper for input tokens ($2.00 vs. $2.50 per 1M tokens) and $10.00 vs. $10.00 for output tokens (1.2x cost advantage). In terms of operational performance, GPT-5 delivers lower response latency at 340ms TTFT (180ms faster than GPT-6.1 Sol) with 105 vs. 135 tok/s throughput. GPT-6.1 Sol leads coding evaluations at 86.4% SWE-bench Verified vs. 74.9% for GPT-5 and 92.4% vs. 85.4% on GPQA Diamond.
GPT-5 vs Grok 3
GPT-5 is 16.7% cheaper for input tokens ($2.50 vs. $3.00 per 1M tokens) and $10.00 vs. $15.00 for output tokens (1.2x cost advantage). In terms of operational performance, GPT-5 delivers lower response latency at 340ms TTFT (280ms faster than Grok 3) with 105 vs. 88 tok/s throughput. GPT-5 leads coding evaluations at 74.9% SWE-bench Verified vs. 65.8% for Grok 3 and 85.4% vs. 84.6% on GPQA Diamond.
GPT-5 vs Grok 3 Mini
Grok 3 Mini is 88.0% cheaper for input tokens ($0.30 vs. $2.50 per 1M tokens) and $1.20 vs. $10.00 for output tokens (8.3x cost advantage). In terms of operational performance, Grok 3 Mini delivers lower response latency at 150ms TTFT (190ms faster than GPT-5) with 160 vs. 105 tok/s throughput. GPT-5 leads coding evaluations at 74.9% SWE-bench Verified vs. 54.0% for Grok 3 Mini and 85.4% vs. 68.0% on GPQA Diamond.
GPT-5 vs Llama 3.3 70B Instruct
Llama 3.3 70B Instruct is 95.2% cheaper for input tokens ($0.12 vs. $2.50 per 1M tokens) and $0.30 vs. $10.00 for output tokens (20.8x cost advantage). In terms of operational performance, Llama 3.3 70B Instruct delivers lower response latency at 145ms TTFT (195ms faster than GPT-5) with 165 vs. 105 tok/s throughput. GPT-5 leads coding evaluations at 74.9% SWE-bench Verified vs. 42.5% for Llama 3.3 70B Instruct and 85.4% vs. 55.6% on GPQA Diamond.
GPT-5 vs Mistral Large 2 (2411)
Mistral Large 2 (2411) is 20.0% cheaper for input tokens ($2.00 vs. $2.50 per 1M tokens) and $6.00 vs. $10.00 for output tokens (1.2x cost advantage). In terms of operational performance, GPT-5 delivers lower response latency at 340ms TTFT (40ms faster than Mistral Large 2 (2411)) with 105 vs. 75 tok/s throughput. GPT-5 leads coding evaluations at 74.9% SWE-bench Verified vs. 45.2% for Mistral Large 2 (2411) and 85.4% vs. 54.2% on GPQA Diamond.
GPT-5 vs o1
GPT-5 is 83.3% cheaper for input tokens ($2.50 vs. $15.00 per 1M tokens) and $10.00 vs. $60.00 for output tokens (6.0x cost advantage). In terms of operational performance, GPT-5 delivers lower response latency at 340ms TTFT (2260ms faster than o1) with 105 vs. 48 tok/s throughput. GPT-5 leads coding evaluations at 74.9% SWE-bench Verified vs. 48.9% for o1 and 85.4% vs. 78.0% on GPQA Diamond.
GPT-5 vs Qwen 3.8 27B
Qwen 3.8 27B is 92.0% cheaper for input tokens ($0.20 vs. $2.50 per 1M tokens) and $0.60 vs. $10.00 for output tokens (12.5x cost advantage). In terms of operational performance, Qwen 3.8 27B delivers lower response latency at 120ms TTFT (220ms faster than GPT-5) with 140 vs. 105 tok/s throughput. GPT-5 leads coding evaluations at 74.9% SWE-bench Verified vs. 58.4% for Qwen 3.8 27B and 85.4% vs. 68.0% on GPQA Diamond.
Frequently Asked Questions & Query Fan-Out
How much does GPT-5 cost per 1M tokens including prompt caching?
GPT-5 costs $2.50 per million input tokens and $10.00 per million output tokens. Cached input reads cost $0.625 per 1M tokens, reducing repeated context costs by 75%.
What are GPT-5's coding and reasoning benchmark scores?
GPT-5 achieves 74.9% on SWE-bench Verified, 95.8% on HumanEval coding, 85.4% on GPQA Diamond graduate-level science, 90.8% on MMLU-Pro, and 98% on MATH-500, earning an overall Inference Value Ratio (IVR) of 59.2/100.
What is the context window, TTFT latency, and token throughput of GPT-5?
GPT-5 supports a 409,600 token context window (128,000 max output tokens) and operates at 340ms average Time to First Token (TTFT) with 105 tokens/sec output streaming throughput.
What are the primary production use cases for GPT-5?
Complex enterprise automation, full-stack software engineering, multimodal agent orchestration, and scientific synthesis.