Google Docs Equation Editor Cheat Sheet: Every Name You Can Type, Its LaTeX Equivalent, and the Ones Docs Rejects

This is the table version of the equation editor post. That post explains how the editor behaves; this one is for looking a name up. If you already know that Space converts a backslash name and that the toolbar is the whole vocabulary, skip to the tables.

The editor opens from Insert → Symbols → Equation (keyboard: Alt+I, Y, E). Inside the box you type \name and press Space; the name becomes a symbol or a structure. Everything below was measured in Google Docs on the web, English UI, on 2026-09-21 and 2026-09-22, in one account on one machine — see §5 for exactly how far each table was checked.

1. Six rules for typing a name

1. Space converts. The space itself is consumed; nothing visible is inserted.
2. Any non-alphanumeric key also converts the pending name — \beta+ gives β followed by +.
3. A digit does not. \epsilon2 stays literal. Put a Space between the name and the digit.
4. Enter, Tab and Shift+Enter do not convert; they leave the box with the text unconverted.
5. There is no autocomplete. Nothing suggests names while you type.
6. An unknown name is not an error. It stays on screen as italic text — no beep, no red mark (table C is the list of names this happens to).

That is the whole input model. Slot navigation (Tab into the next slot of the innermost structure, arrows to walk linearly) is in the earlier post’s §3 and is not repeated here.

2. Table A — the 105 symbol names

Every symbol name on the toolbar is also a LaTeX command with the same meaning, and none of them needs a LaTeX package. So for these five groups the “LaTeX equivalent” column would be the name itself; the tables show the glyph and the verification status instead. Status: ✓ typed as \name + Space and seen to convert; ○ typed in the same session but not screenshot-confirmed (treat as untested); nothing in the table failed.

Greek (40) — all 40 typed, all converted.

Name	Glyph		Name	Glyph		Name	Glyph
\alpha	α ✓		\nu	ν ✓		\varphi	φ ✓
\beta	β ✓		\xi	ξ ✓		\chi	χ ✓
\gamma	γ ✓		\pi	π ✓		\psi	ψ ✓
\delta	δ ✓		\varpi	ϖ ✓		\omega	ω ✓
\epsilon	ϵ ✓		\rho	ρ ✓		\Gamma	Γ ✓
\varepsilon	ε ✓		\varrho	ϱ ✓		\Delta	Δ ✓
\zeta	ζ ✓		\sigma	σ ✓		\Theta	Θ ✓
\eta	η ✓		\varsigma	ς ✓		\Lambda	Λ ✓
\theta	θ ✓		\tau	τ ✓		\Xi	Ξ ✓
\vartheta	ϑ ✓		\upsilon	υ ✓		\Pi	Π ✓
\iota	ι ✓		\phi	ϕ ✓		\Sigma	Σ ✓
\kappa	κ ✓		\Upsilon	Υ ✓		\Phi	Φ ✓
\lambda	λ ✓		\Psi	Ψ ✓		\Omega	Ω ✓
\mu	μ ✓

\epsilon is the lunate ϵ and \varepsilon the ε; \phi is ϕ and \varphi is φ — the LaTeX convention, not the other way round.

Miscellaneous operations (32) — 8 confirmed, 24 typed but unconfirmed.

Name	Glyph		Name	Glyph		Name	Glyph
\times	× ○		\oslash	⊘ ✓		\aleph	ℵ ✓
\div	÷ ○		\otimes	⊗ ○		\Re	ℜ ✓
\cdot	· ○		\odot	⊙ ○		\Im	ℑ ✓
\pm	± ○		\dagger	† ○		\top	⊤ ○
\mp	∓ ✓		\ddagger	‡ ✓		\bot	⊥ ✓
\ast	∗ ✓		\vee	∨ ○		\infty	∞ ○
\star	⋆ ○		\wedge	∧ ○		\partial	∂ ○
\circ	∘ ○		\cap	∩ ○		\forall	∀ ○
\bullet	• ○		\cup	∪ ○		\exists	∃ ○
\oplus	⊕ ○		\neg	¬ ○		\triangle	△ ○
\ominus	⊖ ○		\diamond	⋄ ○

Relations (21) — 9 confirmed, 12 typed but unconfirmed.

Name	Glyph		Name	Glyph		Name	Glyph
\leq	≤ ○		\equiv	≡ ○		\subset	⊂ ○
\geq	≥ ○		\sim	∼ ○		\supset	⊃ ○
\prec	≺ ✓		\simeq	≃ ✓		\subseteq	⊆ ○
\succ	≻ ○		\asymp	≍ ✓		\supseteq	⊇ ○
\preceq	≼ ✓		\approx	≈ ○		\in	∈ ○
\succeq	≽ ✓		\ne	≠ ✓		\ni	∋ ✓
\ll	≪ ✓		\gg	≫ ○		\notin	∉ ✓

\ne is the only not-equal name. \neq, which LaTeX accepts as a synonym, is rejected by Docs (table C).

Arrows (12) — 7 confirmed, 5 typed but unconfirmed. Case matters: \updownarrow is ↕ and \Updownarrow is ⇕, exactly as in LaTeX.

Name	Glyph		Name	Glyph
\leftarrow	← ○		\Leftarrow	⇐ ✓
\rightarrow	→ ✓		\Rightarrow	⇒ ○
\leftrightarrow	↔ ✓		\Leftrightarrow	⇔ ✓
\uparrow	↑ ○		\Uparrow	⇑ ○
\downarrow	↓ ○		\Downarrow	⇓ ✓
\updownarrow	↕ ✓		\Updownarrow	⇕ ✓

\to, LaTeX’s usual short form of \rightarrow, is rejected (table C).

3. Table B — the 20 structure names, and what they are in LaTeX

The Math operations dropdown is where the names stop being LaTeX. Thirteen of the twenty are LaTeX commands or close to them; seven are Google’s own. Each one creates a structure with empty slots (Tab moves to the next slot). The LaTeX column is what the same structure would be written as in LaTeX, with a, b, c standing for the slot contents. Status: ✓ typed as \name + Space and confirmed; T inserted from the toolbar and confirmed, not typed; — not tried by anyone yet.

Docs name	Builds	LaTeX equivalent	Status
\frac	fraction, two slots (caret lands in the numerator)	\frac{a}{b}	✓
\sqrt	square root, one slot	\sqrt{a}	✓
\rootof	root with an index slot and a radicand slot	\sqrt[a]{b}	✓
\superscript	base and raised slot	{a}^{b}	—
\subscript	base and lowered slot	{a}_{b}	—
\subsuperscript	base, lowered and raised slots — three slots	{a}_{b}^{c}	✓
\overline	bar over one slot	\overline{a}	✓
\widehat	hat over one slot	\widehat{a}	✓
\sumab	∑ with lower and upper limit slots, body typed after	\sum_{a}^{b}	—
\prodab	∏ with limit slots	\prod_{a}^{b}	—
\coprodab	∐ with limit slots	\coprod_{a}^{b}	✓
\bigcapab	⋂ with limit slots	\bigcap_{a}^{b}	T
\bigcupab	⋃ with limit slots	\bigcup_{a}^{b}	T
\intab	∫ with limit slots	\int_{a}^{b}	✓
\ointab	∮ with limit slots	\oint_{a}^{b}	—
\limab	lim with two slots underneath and an arrow drawn between them	\lim_{a \to b}	✓
\rbracelr	auto-sized ( ) around one slot	\left( a \right)	✓
\sbracelr	auto-sized [ ] around one slot	\left[ a \right]	T
\bracelr	auto-sized { } around one slot	\left\{ a \right\}	T
\abs	auto-sized vertical bars around one slot	\left\| a \right\| (single bars)	✓

Three things the table cannot show:

• Typed shorthand is different from the palette name. Typing \lim + Space gives a lim with one slot underneath (\lim_{a}); the palette’s \limab gives two with an arrow. Typing \int, \oint or \sum + Space gives the operator with limit slots, the same shape as \intab, \ointab, \sumab. ^ and _ create a raised or lowered slot immediately, no Space needed.
• The bracket structures are real auto-sizing brackets — the ( ) built by \rbracelr grows to fit a fraction inside it, exactly like \left( … \right) in LaTeX. Typed ( and ) are plain characters and stay character-sized.
• The seven Google-only names — \rootof’s LaTeX cousin exists, but \subsuperscript, \bigcapab, \bigcupab, \coprodab, \rbracelr, \sbracelr, \bracelr (and the -ab forms generally) mean nothing to LaTeX. If you ever retype a document’s equations in LaTeX, those are the names to translate with the column above.

4. Table C — names Docs rejects, and what to type instead

These stay on screen as italic text. No error, no underline (three of them get a spell-check squiggle), and the equation looks fine at a glance — which is why they are worth listing.

You typed	Because in LaTeX it is…	In the Docs editor
\neq	the usual not-equal	type \ne
\to	short for \rightarrow	type \rightarrow
\left( \right) \bigl(	auto-sized brackets	use the toolbar’s ( ) [ ] { } | | entries, or type \rbracelr \sbracelr \bracelr \abs
\{ \}	literal braces	\bracelr for a brace pair; a typed { is a plain character (why that matters, §9)
\iint	double integral	two \int in a row; there is no single ∬
\begin{matrix} \matrix \pmatrix	matrices	cannot be built — see §6
\cases	piecewise function	cannot be built — there is no line break inside a box
\text \mathbf	text and bold inside math	cannot be built — the box has one font
\langle \rangle \lfloor	angle brackets, floor	not on the palette; no equivalent found

The pattern: if a name is not on the toolbar, Docs does not know it. The palette is not a subset of what you can type — it is the whole list, plus a handful of function names (§5).

5. How far each table was verified

The claim “every palette name is typeable” started as a spot check, so here is the exact state.

• 79 of the 125 names have been tested and every one converted: 75 typed as \name + Space, 4 inserted from the toolbar only. Greek 40/40; miscellaneous 8/32; relations 9/21; arrows 7/12; structures 11/20 typed plus 4 toolbar-only. Exceptions found: none.
• 41 names were typed but not screenshot-confirmed (the rest of miscellaneous, relations and arrows) — a capture failed partway through the session. They are marked ○ above and count as untested, not as failures.
• 5 structure names have not been tried by anyone (\superscript, \subscript, \prodab, \sumab, \ointab) — in practice people type ^, _, \prod, \sum, \oint, and those work.
• Function names are not on any dropdown, yet \sin, \cos and \log + Space convert to upright function names. \vec, \hat, \bar and \binom also convert. The full set of names that work without being on the palette has not been measured; assume nothing beyond these.
• One account, one machine, English UI, September 2026. No Docs version number is exposed to check against.

What the editor stores for these names is a separate question with a separate measurement: the symbol names come back from Apps Script as EQUATION_SYMBOL nodes carrying the same code, the twenty structures as EQUATION_FUNCTION nodes, and \sin as a symbol rather than three letters — that post has the trees.

6. When the list runs out

Tables A and B are the vocabulary, and table C is its edge. Matrices, piecewise functions, aligned lines, text inside math and anything with \left are past the edge, and no name typed into the box will bring them in.

The way around is the same one the earlier post ends on: write those constructs as LaTeX text in the paragraph — $$\begin{pmatrix} a & b \\ c & d \end{pmatrix}$$ — and let LaTeXFlow for Google Docs render the text into an equation image in place. The image is not a native equation object, but it is a matrix, which the native editor will never be. Installing it and converting a first equation takes about two minutes.

---

• How the editor behaves while you type, with screenshots → The Google Docs Equation Editor: Typing It Fast, and the Four Things It Cannot Build
• What each of these names becomes in the document’s structure → What Google Docs’ Equation Editor Actually Stores