Grok 4.3 vs GLM 5

Pricing math

Grok 4.3 lists at $1.25/M input and $2.50/M output from xAI; on OpenKey that's $1.25 x 1.03 = $1.2875/M input and $2.50 x 1.03 = $2.575/M output. GLM 5 lists at $0.60/M input and $1.92/M output; on OpenKey that's $0.60 x 1.03 = $0.618/M input and $1.92 x 1.03 = $1.9776/M output. Grok 4.3's input price is 2.08x GLM 5's. Run the numbers on a real workload — 10M input tokens plus 2M output tokens — and Grok 4.3 costs $17.50 versus GLM 5's $9.84. That's a fixed gap you pay on every batch job, not a one-time difference. Both models are billed through OpenKey with a single key and the same flat 3% fee on top of provider list price, so there's no platform markup difference to factor in — the price gap is entirely between xAI and Z.ai's own list prices.

Agentic and coding benchmarks

On the Design Arena categories both models are scored in, GLM 5 leads across the board: androidnative (elo 1244, rank 6, 62% win rate) versus Grok 4.3 (elo 1074, rank 23, 28.6% win rate); fullstack (GLM 5 elo 1190 rank 13 vs Grok 4.3 elo 1072 rank 26); godotgamedev (GLM 5 elo 1237 rank 3 vs Grok 4.3 elo 1134 rank 16); mobileapps (GLM 5 elo 1222 rank 10 vs Grok 4.3 elo 1146 rank 26). The pattern repeats in model-level categories like codecategories (GLM 5 elo 1295 rank 16 vs Grok 4.3 elo 1243 rank 31) and gamedev (GLM 5 elo 1299 rank 15 vs Grok 4.3 elo 1242 rank 34). Grok 4.3 does hold its own on Artificial Analysis metrics not measured for GLM 5: intelligence index 37.6, coding index 42.2, agentic index 24.1 — but there's no matching GLM 5 score to compare directly.

Context and modality

Grok 4.3 supports 1,000,000 tokens of context against GLM 5's 202,752 — a 4.93x ratio. If you're feeding in full codebases, long transcripts, or multi-document retrieval sets, Grok 4.3 has meaningfully more room before you need to chunk or summarize. Grok 4.3 also accepts image and file inputs alongside text; GLM 5 is text-in, text-out only. If your pipeline needs to read screenshots, PDFs, or diagrams, GLM 5 is out regardless of its price or benchmark advantages. Neither model publishes a max completion token cap in this data, and neither has a listed knowledge cutoff date.

When to pick each

Pick Grok 4.3 when the task involves images, files, or documents beyond 202K tokens — the 1M context window and multimodal input are the whole reason to pay more. Pick GLM 5 for pure coding and agent workloads at scale: it wins on every shared Design Arena benchmark and costs $9.84 versus $17.50 on the same 10M/2M workload, a difference that compounds fast in CI pipelines or batch agent runs. Both are available on OpenKey under one key with the standard 3% fee, so switching between them for different jobs doesn't mean managing two sets of credentials.

Questions

Which is cheaper, Grok 4.3 or GLM 5?

GLM 5. On OpenKey it's $0.618/M input and $1.9776/M output versus Grok 4.3's $1.2875/M input and $2.575/M output — Grok 4.3's input price is 2.08x GLM 5's. On a 10M-input/2M-output workload, that's $9.84 for GLM 5 versus $17.50 for Grok 4.3.

Which model has a bigger context window?

Grok 4.3, by a wide margin: 1,000,000 tokens versus GLM 5's 202,752, a ratio of 4.93x. If your workload needs more than 200K tokens of context in a single call, GLM 5 can't handle it and Grok 4.3 is the only option here.

Can GLM 5 process images?

No. GLM 5's modality is text-to-text only. Grok 4.3 accepts text, image, and file inputs and outputs text, so any workload involving screenshots, PDFs, or diagrams has to go to Grok 4.3.

Which model wins on coding benchmarks?

GLM 5, on every shared Design Arena category. In codecategories GLM 5 scores elo 1295 (rank 16) versus Grok 4.3's elo 1243 (rank 31). In fullstack GLM 5 hits elo 1190 (rank 13) against Grok 4.3's elo 1072 (rank 26).